"Adapting existing free software projects" is not that common as a business.
And you are right: the actual developers of the project are the first (and
sometimes the only) ones to make money in this way or by implementing desired
new features.
I believe more common (small) businesses do consultancy (what software to
choose, how to migrate, etc.), integration (how to configure, how to make the
applications talk to each other, etc.), assistance (updates, fixing problems
in brief delays, etc.), formation (how to use LibreOffice in an efficient
way, what are the new features of the application that has just been updated,
etc.), etc.
I agree the FreeRunner somewhat failed. Sure, the phone was not perfect..
especially the bugs preventing to give/receive some calls! However, don't you
think a little more financial involvement from OpenMoko would probably have
bear fruits? At least I do not see any reason why this business model would
be doomed to failure.
I also agree that Mozilla's business model is quite peculiar (although very
effective). I would actually rather prefer that Mozilla makes a deal with a
search engine such as DuckDuckGo or Ixquick, which are more respectful of the
privacy of their users. On the other hand, Mozilla dealing with Google only
has advantages for Trisquel's users (who unfortunately are a minority): this
brings piles of money to Mozilla, hence many paid developers making a great
browser... we use without the Google search!
