I just can't think of any other reason why they're both so infamously bad.
If the site works for the 95% of ad dollars who use Chrome on their phones, it works as intended.
Excellent point. If any change on the site caused ad views or clicks to drop by >1%, automated tests probably shoot flares up to the exec level. Whereas the vast majority of users on FF are blocking ads anyway, so their lagging performance probably barely registers on those tests.
While YT probably still has non-ad automated performance tests, in the case of a non-Chrome, non-Apple desktop browser, those tests probably run every odd Thursday and regressions send a toast message to an intern. :-)
The software engineering show isn't being run by engineers, middle-managers or even MBAs anymore.
It's run by product people who treat it like a mini dictatorship.
So releases like this have to go out, no matter the consequence.