They probably all use macs though, why not develop first for something you understand well and can immediately put to use yourself. It probably started as a fairly experimental product, now that they have a version out and a clearer idea they could more easily tackle a windows version.
You're not describing a bubble, "making something I want" isn't some outlandish absurd thing for developers and designers do. Bubble is rapidly outpacing pivot for most misused word.
He meant bubble as in 'echo-chamber', and in that sense of the word it isn't overused.
It's not outlandish or absurd to design something you might want, but it is a bit silly to ignore a larger market with the same underlying need.
They're not ignoring a larger market, they're ignoring a smaller one:
> Beyond that, one-third of our traffic uses Macs. It's not a small market for us.
I feel like I must be missing something here; how is the 2/3rds that don't use Mac OS the smaller market?
And as others have pointed out, the market doesn't just consist of GitHub users, or even Git users. In this case, it's developers who use an RCS and need a good desktop application to go with it.
In any case the point may be moot; judging by the rest of the comment you quoted they're not ignoring other platforms at all but rather just using Mac OS as a first step, which I can understand.
Twitter has a Mac app because Loren Brichter wrote a fantastic one and Twitter bought it (and Loren too). I'm not particularly aware of the state of Twitter on Windows, but I find it plausible that there is no single excellent Twitter app that Twitter wants to buy (or if there is an excellent one, maybe they don't want to sell).
They are the latter. Welcome to the San Francisco bubble, where you think everyone uses iPhones and Macs, because in your world everyone does. It's the same reason Twitter has a Mac app but not a Windows app. They build these things for themselves (and arguably they should).