I just pulled the trigger on an Ultimaker a few weeks ago after seriously considering a Thing-O-Matic from Makerbot. I liked the fact that the build surface only moves along Z and isn't sliding around all over the place. It just seems to me a design with a moving toolhead with involved forces being predictable is better than a moving build surface of slowly increasing mass sliding around, and that it would be more useful in using more of the printer's volume as opposed to TOM that requires enough space on the right for left-justified deposits, etc etc.
Turns out there are quite a few different hobby-grade 3d printers out there.
http://thefutureis3d.com/ can send you an assembled one that's 24" x 24" x 13.5"... crazy huge (about 125,000 cm^3 if I don't suck at stubbornly-refusing-to-write-digits-down math)!
So, point being that while the idea that Makerbot's has serious money to invest in R&D is good, but, from a business standpoint, it's land-grab time. Give a bunch away to schools and prominent bloggers, get puff pieces in magazines, spend a little for advertising, you know, the stuff to firmly embed themselves in the public's psyche, almost to the point where if people think about 3d printer they would immediately think of MakerBot instead of RepRap or Up (FDM designs) or even think of that before, say, ZCorp (powder based). The design has a lot of catching up to do.
If they don't, the first one past the post I think will be the first design that has enough sensors (humidity, ambient temperature, filament chemistry detection) and software wizardry (pathing, automated support structures) to remove a lot of the guesswork and experimentation that everyone with these devices tend to go through now. But to roll the dice that your R&D will be first to do that and do it well? Hmm...