[acdha]

I think this is correct and would add that the YouTube example highlights that it’s a management decision, not just an inexcusable lack of testing. I’d definitely believe that developers tune what they use first, but once a known problem is identified someone had to decide how to prioritize it.

In many ways, I’d treat this like mandatory banking & investment separations where a cost of being a browser developer should be that you can’t ignore things other shops can. Vimeo could decide not to fix a performance regression affecting Firefox but YouTube should be required to fix it within 60 days, and if they don’t like that they can split it into an independent company which wouldn’t have that constant conflict of interest concern.