Well, I certainly agree with the point that the Internet has intensified the success of open source. But to avoid the either/or kind of argument, I think a more nuanced position advocating the co-construction of open source technologies might be better (i.e. both structure *and* agency were important - technology and users/developers).
In any case, I think in trying to find a kind of empirical 'community' is barking up the wrong tree. The open-source community could function more as an 'imagined community' (rather like the way that Benedict Andersen famously argued) so that it is the belief it exists that creates the feeling of belonging to a community (which may or may not be empirically verified). This could then feed into an eventual 'real' community as people start to map it, draw up technology support systems (like fora and wikis) and start to develop a set of common values. This raises the interesting question as to what extent the open-source community is building itself ;-)
I think you should also be careful about asserting the 'ideology' of free software without a critical attention to the underlying ideology of the open-source advocates (even being apolitical is a political position, afterall). I would suggest looking at
Which is an attempt to map these divisions and try to understand how the technical can also contain values (and by definition politics).
However, I do think you have made some important arguments, and raising them for debate is always to be applauded.