I share most of your thoughts, but: where do you get the initial money from? Even if you gut all the software development costs by creating a "community phone", you still haven't gotten past either A, the financial hardware requirements, and B, market entry costs/barriers/difficulties.

Frak, if it was that easy, I'd pay for a domain right now, take out a nice loan (if any bank would give me a big enough one as a jobless college student) and start the process. Only you still have to make contact with all the relevant hardware manufacturers, for all the devices/hardware you buy by the bulk that require open-source drivers that MeeGo doesn't have written yet, you have to get driver writers to make them. Personally, I also dislike capacitive with a passion, and Stantum is the only resistive multitouch available now, so I'd say license stantum screens and then see if the right MeeGo drivers already exist, and if they don't, you have to make your own... and you damn well better be ready for someone somewhere to sue your *** with their multi-million-dollar legal teams. And then there's the factory connections. I don't even know how you'd start on that - is there a place you can even go to look up what factories operate where, what devices/etc they can make? Someone has to assemble all the parts, and I sure as hell don't expect the community members to do it by hand... Actually, if I were to ever have a say in a project like this, I would push strongly for future upgrade-ability, so I would push manufacturers of the parts to make them at least somewhat plug-and-play-ish, but honestly, you can't take that very far on a phone without significantly impacting i/o speeds and sheer volume/size of the device. Either way, someone has to initially pay the tens of thousands of dollars to bulk-order SoCs, WiFi cards, penta-band GSM (and maybe CDMA) antennas (I know CDMA sucks in some ways, but so long as it exists, I believe it would be nice to include both antenna types in one device, and that's done way too rarely), etc, etc. God forbid you decide to give the users options, of, say, capacitive vs. multitouch resistive screens, or hardware keyboard vs. no keyboard variants. And then once you've invested that money, that someone isn't going to want to send these parts out in small quantities to 'community' members with the hope that they assemble them correctly. So you have to make some connections with factories off in China/Taiwan/etc.

I mean, I can already imagine that assuming the 'community' as a whole is so passionate about the project to make the drivers and software kinks happen, the hardware costs alone should be reasonably easy to turn a minor profit on. Hell, even 10-30$ of profit, to keep the site going and pay for the devices that are given out to the 'testers', early on. But keep in mind anyone maintaining this would have either investors or lenders to massively pay off, as well as probably lost job time just to make it happen.

And while all this is going on, the 'community' is busy doing one of three things:
1: The loyal volunteers keep pushing updates/patches/goodness, spreading the word and so on - but those numbers dwindle because there's no way in hell a completely new group can make the above happen fast - just look at Notion Ink. It took them like two years to go from prototype to putting Adams into buyer's hands, and after all the hype, only the first pre-order-ers have gotten their Adams - minus a small portion who got even more delays.
2: The majority of the public goes through waves of excitement, anger, entitlement, and being distracted by something new and shiny.
3: The tech bloggers and sites mostly fling around words like vaporware and the like. This is partially mitigated by having open and transparent development, but to even get that off without a hitch, you still need the expertise with managing projects - and the expertise of not just running a bugzilla and code repository and so on, but for true openness, posting numbers of testing devices being sent, possibly keeping an honest account of donations and all the costs incurred, and then running things like hardware bug/problem QA.

Hell, it's doable. But someone has to be dedicated enough to make it happen. Someone, right now, in the 'community' has to be able to acquire the start-up funds to make all this happen, and the willingness to drop a significant chunk of their life, to do so. I am mid-university-education, I have a partner/loved-one who's safety and financial security I'm not willing to completely hinge on whether or not something like this is pulled off. I have a pretty strong faith that IF this was seen through to the end, it would make money, and could become genuinely game changing... Alternatively, if it fails, because in the end something goes wrong before you can start selling, and you can't invest any more money, but there's tens of thousands of dollars already invested in hardware procurement/production costs... you're very f|_|cked.