It would also depend on the type of home you are building. We are building a country triple dormer where part of the ceilings on the upstairs bedrooms are sloped with the roof line. In our instance, we chose to go with foam though much more costly. We had a bonus room with same type roof slopes and it was always warmer in summer and cooler in winter. It had fiberglass. Do you have a vaulted ceiling anywhere? Cellulose will eventually settle over time as well.
Your exterior material will also affect it. Brick and stone help more than siding.

In our current build, the only places that got fiberglass were detached garage and in the walls between bedrooms. Everything else is foam. We are in Arkansas.

I would think the savings in heat from going from 16 to 19 in Birmingham on the walls would be like $30 for a 3000 sq ft house and that is being really generous. My walls on a 4000 sq ft house in NC cost $200 a year. That is for r-13. Upgrading to R-19 would save about $60 a year. A much bigger upgrade than you are considering in a warmer climate so that is where my $30 comes from. So you might be talking a 200 year payback.

The cooling cost for my walls is about $50 a year. Even if yours were twice mine, you'd be looking at another $10 in savings. So you might be getting down to a 150 year payback.

Just to give you something to think about. Cooling is a really minor issue with wall insulation. Even 95 outside and 75 inside is a paltry 20 degree difference. Factor in nighttime temps of 75 with zero degree difference and your average delta T is only 10 degrees.

A cheaper option might be more fiberglass in the attic - you can easily go to r-40 without a problem

Thanks, bh401 and David.

bh401 -- no vaulted ceilings. We're doing three sides brick and one side Hardie plank.

David -- thanks for the estimates. Even with my new exercise program 150 years may be pushing it!

I assume you were doing .5 XPS over osb or something like that? Otherwise .5 is way to thin and will bulge with any pressure whatsoever from a dense packed product. batts would ok probably.
If you are after air sealing, .5 XPS and batts will not get you there. While high r and reduce thermal bridging is not really a goal for your area, air sealing for conditioning would be. I would recommend looking at different exterior product that will get you better sealed up for simialr (possibly reduced) costs. Cellulose is not an air sealer, but rather the density of cellulose vs fiberglass is an air buffer. But certainly not at 3.5". You need to be 8"+ before it really starts to reduce air movement through, packed pretty good at 2 lbs min. I would not fret about losing the air infiltration battle vs cellulose and batts with 3.5".
For the exterior, I would not use the exterior XPS and put that money toward air tight osb approach and caulking. Using Hubber ZIP sheathing will give you a tightly sealed wall and water drainage plane. Double up on cutting down infiltration and caulk each stud prior to placing the sheathing. Caulk the sill and head plates to each other and the floor, caulk the headers to the framing, etc. at a few bucks a tube and a few seconds to apply it, its is the cheapest insurance toward reducing infiltration possible.
Air tight osb is similar to ZIP, but the sheathing is about half the price. However seams need to be primed and taped off, and then you need a moisture drainage plan, so a building wrap of some sort also needs to be applied and taped. By the time you add the products and time, you are past the price of ZIP sheathing.
After you have your shell taped off, caulked and sealed, then your insulation type for a 3.5" cavity becomes less crucial for sure. Batts would be just fine for this application. The important thing to remember is neither batts or cellulose will reduce your air infiltration....
also, if you have not done so, discuss with your HVAC supplier the need for an ERV as part of your system. You will need one for any sealed home.

I assume you were doing .5 XPS over osb or something like that? Otherwise .5 is way to thin and will bulge with any pressure whatsoever from a dense packed product. batts would ok probably.
If you are after air sealing, .5 XPS and batts will not get you there. While high r and reduce thermal bridging is not really a goal for your area, air sealing for conditioning would be. I would recommend looking at different exterior product that will get you better sealed up for simialr (possibly reduced) costs. Cellulose is not an air sealer, but rather the density of cellulose vs fiberglass is an air buffer. But certainly not at 3.5". You need to be 8"+ before it really starts to reduce air movement through, packed pretty good at 2 lbs min. I would not fret about losing the air infiltration battle vs cellulose and batts with 3.5".
For the exterior, I would not use the exterior XPS and put that money toward air tight osb approach and caulking. Using Hubber ZIP sheathing will give you a tightly sealed wall and water drainage plane. Double up on cutting down infiltration and caulk each stud prior to placing the sheathing. Caulk the sill and head plates to each other and the floor, caulk the headers to the framing, etc. at a few bucks a tube and a few seconds to apply it, its is the cheapest insurance toward reducing infiltration possible.
Air tight osb is similar to ZIP, but the sheathing is about half the price. However seams need to be primed and taped off, and then you need a moisture drainage plan, so a building wrap of some sort also needs to be applied and taped. By the time you add the products and time, you are past the price of ZIP sheathing.
After you have your shell taped off, caulked and sealed, then your insulation type for a 3.5" cavity becomes less crucial for sure. Batts would be just fine for this application. The important thing to remember is neither batts or cellulose will reduce your air infiltration....
also, if you have not done so, discuss with your HVAC supplier the need for an ERV as part of your system. You will need one for any sealed home.

Thanks, lzerarc. The lack of air sealing with cellulose was a surprise to me. We're going to discuss the Huber ZIP sheathing with our architect. In fact, I've just sent the e-mail.

Jeff - it was a surprise to me that you thought cellulose was air sealing. Lezarc knows what he is talking about.

I'm always surprised how expensive insulation is for traditional builds. Shaving off $6000 by going from cellulose to fiberglass??? How big is the house?

My house is full masonry (not just veneer but load bearing) and wrapped with 4" of Type-I EPS all around for about $5000 total (3600sf wall area).

jrldh -- the house will be a two-storey with a total of about 3100 square feet.

David -- I believe that lzerarc knows what he is talking about. I think that shows by my incorporating his feedback into our discussion with our architect. I can assure you that my expressing surprise was not intended to be an expression of skepticism about his response.

Jeff:
I re-read your post. One thing that jumped out at me was your estimated r numbers. Who gave you these? Dense pack cellulose is around 3.8-4 per inch r, so x3.5" is right around an r-13, plus 2.5 r for the .5" XPS (r5/ inch) giving you around a r-16 CAVITY. Your studs with the XPS will be somewhere around an r 6. Take all of this into account, and your clear wall with cellulose will be around r-12-13 range (without doing the math, your .5 XPS helps you slightly for a thermal break on the framing), less for batts. However factor in infiltration, and you have about an r-0 (truly) no matter which system.
For my projects I like to focus on u values, and not r values. The companies that product the stuff (dow, Owens, etc) have done a great job at marketing r value, even though it tells a very small part of the story when it comes to the walls performance. I would say 80% of the people I talk to see an r-21 stamped on the fiberglass batts packaging and think they are going to have a really energy efficient house. This is exactly the marketing game they play, and want you to believe. They are not all evil of course, as they make some super products to use in sealing and high efficient construction, its just a lot are not readily available off the shell nor does anyone do the research on their site to educate themselves. U values focus on the wall assembly. Get the u values down, r will go up.
But like I said, for your area, kill the air infiltration, and that will make a huge starting difference. I would also be curious to hear the feedback your architect gives for ZIP sheathing.

The lack of air sealing with cellulose was a surprise to me.

And probably, too, to Paul Fisette, a prominent Professor of Building Studies and Technology at the University of Massachusetts., who recommends cellullose for that very reason. But, of course, its air sealing qualities are not a substitute for careful air sealing with caulking and foam. (I would be curious to see a report that filling a wall with cellulose insulation will bulge sheathing off the house framing.)

Indeed, spray foam is a favourite of Building Science Corp. in its model warm climate homes, such as this one.

lzerarc -- I took the R-values from the drawings that were provided by our architect. I can't remember the specifics right now, but when I get home this afternoon I'll post them.

worthy -- Thanks for the links. (The first link appears to be broken, but it was easy to find the document from the title.) I really like the Building Science site.

worthy
I read your referenced article. We are basically saying the same thing. I think the word "sealer" is the one we need to look at. Cellulose is not an air sealer, and I would hope we can agree on that (nor does Paul F claim its a sealer). Rather the density buffers or slows down the transfer of air through it better then fiberglass, which is also what I stated.
I will say, for my area, I always spec or recommend cellulose. His upcharge price does amaze me however. Cellulose, at least in our area, is cheaper then fiberglass to purchase, slightly more to install, so it typically ends up being a very slight upcharge.

For the bulging...there is no report. However .5" XPS is from field experience. OSB sheathing, no. But .5" XPS- yes. I have seen 1" XPS bulge slightly from cellulose. (not overly visible with the eye, but set a 4' level across the exterior face...it "rocks" over the cavity location).

lzerarc: Here's how our architect determined the R-value for the cellulose option.

Brick: .2*3"= .6 + .25 (I don't know what accounts for that extra "+.25" here.)
Air space: 1"= 1.01
XPS: .5*5 = 2.5
Cellulose: 4*3.5" = 14
Gyp. bd.: .9*.5" = .45
Inside surface: = .68 (I don't know what the inside surface refers to unless it's paint.)

Total: .6 + .25 + 1.01 + 2.5 + 14 + .45 + .68 = 19.49.

(begin rant)

You boys are talking dense-packed cellulose for the walls, right? I read about "settling". Settling will be a factor for loose-blown in the attic, but not for cells properly blown into framing cavities.

If I were to dense-pack a wall cavity that only had 1/2" XPS, it'd blow the XPS off the wall. XPS over OSB or ply? No worries there.

Cellulose, while not being an air sealer in a 2x4 cavity, does knit together and is superior to FG batts in slowing air movement in or through a cavity. Especially since FG batts are never really correctly installed. (continue rant)

Cellulose on an attic floor is more effective than FG, especially for preventing summer heat gain into the living space through an unfinished attic.

We have "laboratory R-values" and we have "effective R-values". A dense-packed cellulose installation will effectively outperform a fiberglass installation in my climate (CT) any day. And any night.

Sound? Cells win.
Air? Cells win.
Thermal? Cells win.

So yeah, I favor cells over FG. lol

But what kills me is...a $6k premium for cells? Wow.

(end rant) lol

jeff
I see. I was looking at insulative values only, which I typially only compare (its easier that way!) but yes, to get the entire wall stack up (through the cavity as he is showing) that would be correct. Your clear wall will be less, as well as your performance of your wall (taking into account framing, penetrations, openings, etc)
Also, for dense packing, I would even use 3.8-4 for the cellulose. 3.5 is more for loose blown, as in an attic (which you DO NOT want to use in your wall, as it will settle).
A lot of talk and focus is and has been on the walls. Your house as 6 sides...and they must not be ignored either as they back up the entire shell. Hopefully you have around an r-40ish in your attic? Slab on grade I assume? I will admit, I am not overly familiar with below slab applications for your climate, but in ours I typically put 1.5-2" of XPS under the entire slab. (but then again we freeze at 3-3.5' below grade).
Also do not forget about your windows. Windows are the largest source of energy loss in a building most of the time....but designed/detailed correctly, they can also be a massive energy saver.