> IMHO the FreedomBox project needs more differenciation features...
No, if they want to achieve their stated goals they need less. In that they need to create their stuff in such a way that they can be widely deployed to the point they become expected. That probably won't happen if it is only available as a special purpose hardware device. So they need to be thinking more along the lines of a set of portable packages along with a reference system image.
Most ISPs will be happy to rent you a WiFi router/AP but will allow you to connect your own equipment. Being outside the NAT would eliminate half of the problems with the Freedom Box. Yes there are places where you can't replace the outside facing hardware. Yes there will be people wanting to run inside a NAT... and if we assume IPv6 isn't going to happen in the next decade there will be a lot of those in places where IPv4 address exhaustion is a problem. But the best way to evolve software is when it is in active use. Get it running NOW and adding features will happen faster. Stop building a perfect Cathedral and raise a flag out in the bazaar as soon as you have something that will inspire.
I'm running OpenWrt now. Dual band abgn with an untainted kernel. The hardware is available, you just have to shop careful. Instead of chasing what could be, deploy on what is available now. If asking users to reflash is an issue, sell preflashed units and raise a little extra scratch to fund development. I'd bet you could do that and still be less expensive than a plug.