"The upside is that they should have a bit more low end than something like the Fi IB318."
they won't have more low end than the ib318. that driver has something like 30mm xmax. since both drivers are infinite baffle variety, very little power is required to push them to their excursion limits. so on the low end, they are both excursion limited (as compared with power limited higher up in frequency).
that said, the manifolds to reduce vibration are a good idea unless you can build that front wall with I-beams (H).
in theory, with the front wall loaded sufficiently, side to side and top to bottom room modes are all cancelled, leaving you with just the front/back wall modes, which with one row of seating are all the same, so the entire row can be eq'd identically and more or less to perfection. with two rows, there is a little tradeoff, but still should be really good.
paulw's octagon (member pnw) is definitely worth another visit. his aren't arranged optimally, but like you will have, there are so many, that optimally doesn't matter so much.
how much space do you have to work with up there? 20 cf per driver models quite nicely as a large sealed/small i.b. with that driver.
and just in case it isn't clear, I think you made a great choice. if all up front with your mains, you will be able to crossover where you want to your mains with headroom out the wazoo...114db 1w/1driver/1m sensitivity in 2pi space, but yours will be much higher because it is effectively creating a plane wave tube in your theater. somebody might be able to help me out on this one for the sensitivity gain, it is at least +6, but I think more, so you will be over 120db 1w/driver/1m and because of the plane wave nature...no rolloff in spl with distance...so 120db+ 1w/driver in your face.
**btw 1w/1driver/1m is my homemade method for reporting total system sensitivity. I think that gives a reasonable comparison where drivers are in the same ballyard for power handling.