It's not selective: it's specifically about the desktop space. I've mentioned that they did fine on the server.
So, IIS and xBSD etc is out of scope regarding what I've said. I also mentioned explicity community/bazaar projects, in the original FOSS spirit of the late nineties that was said to conquer the world.
As for Firefox and Chrome, both are based from code by big (at least at the time) corporations (Netscape, and Apple, and later Google).
If the latter didn't get involved with Webkit, I don't see us using KHTML 2016 today. As for the former, Mozilla (which begat Firefox) also started with big commercial backing with full time employees from Netscape for several years (until the company burned to the ground), and until today works with mostly paid Mozilla staff, not random community people. It was able to do that based on inheriting some of the momentum of Netscape (which at one point was the number 1 browser), and, based on that, to ask for millions from Google etc to have it as the default search engine. Without Netscape's legacy code, 2+ years of refactoring/engineering work, and initial user base (at the time of the transition) it wouldn't have gone far.
In other words, yeah, open source can do a great job on the desktop too, if it starts as a proprietary code donation and has huge funding. But that "open source" is just a licensing model, so of course it won't be any different in any characteristic to proprietary projects, apart from the licensing -- the CLR released by Microsoft is also "open source" under this metrics.
My comment was about the FOSS bazaar model of volunteers from all over the world, the one touted by the Gnome, etc teams, -- which didn't fare as well on the desktop space.