You've got to admire Richard

> That assertion is just false on its face - we lose many developers to Windows and Mac OS X
due to the easy availability of GNU software on those platforms. (we can argue about that if
you disagree with it.)

Rather than an invitation to argue about it, which is kind of pointless, I'd prefer to see
some supporting data for your assertion. Given your position, I'm sure that if anyone can
produce such data, you can.

However, "people are unlikely to do X because of Y" is a rather different kind of statement
than just "people are unlikely to do X" - closer to speculation than assertion. Moreover, it's
not an absolute statement - "are unlikely to" isn't a synonym for "won't", and the production
of even a significant minority who do X, and do X only because of Y, does not affect the
acceptability of the statement.

On the other hand, an assertion, as you have made, that "people are doing X right now" *is* a
factual claim, and *does* need to be backed up with evidence. And if you wish it to counter
Stallman's statement, you'll also have to demonstrate that the presence of GNU tools on a
non-free OS makes a user likely (ie. over 50% chance) to switch to that OS in itself, rather
than merely reducing resistance to switching for other reasons.