The foundation of this idea is very promising and has the potential of making it far in the current maker and indie culture.

There are a few points to make which I think are critical for a wider adoption to be attained (wider than openmoko has had until now).

The device has to actually be usable. Nowadays that means at least 1GB of RAM. I would like to echo Estel:

700€+ is a lot of money for something that's a hobby thing and not usable for day-to-day use like the N900 still kind-of is. For me the price I'm willing to pay for something like this would jump from 0 to "as much as it takes" the minute it's actually usable and is relatively future proof. Which leads me to my next point.

If something like this is done, it should be done right. Openness is important, but what biketool is saying about including a UHF/VHF transmitter struck a cord. Especially in todays context of cellular networks being co-opted by governments or just for more robust communication a fall-back alternative for creating ad-hoc networks would be great.

Finally I'd like to summarize that for this to be worthwhile it has to do more than just equal the N900. I can still buy an N900. It might not have great support, but it's "cheap" compared to what a "Neo900" would cost. So if this is done, it has to be done right. Which means day-to-day usable in terms of hardware and future proof in general.

If those conditions were fulfilled I would be amongst those who were willing to pay an arm and a leg for this, not to mention contributing to community support where possible.