If it's gone unfixed for 2 years, should we be thinking there's a hardware issue that Apple is ignoring due to their cost to fix it?
On the other hand...if ONLY Safari can't use YouTube, it really seems like a Safari issue.
I'm wary of professing anything too dramatic as lots of these conversations are derailed by hysterical rants and accusations repeated more than the one time needed to make a point. However, with all of that said, I'm still not ruling out a widespread hardware issue. Primarily because Apple is undoubtedly aware that this is affecting a large customer base. Not large enough to make this a widespread story but large enough that plenty of 'me too' replies immediately pop up whenever this subject is raised as a technical question on forums with many saying they've contacted technical support or filed a bug report/feedback. So Apple is aware that this issue exists and probably has an estimate of how wide an issue it is from a better understanding of its minor hardware parts. It's hardly just shrugged its shoulders, no matter what frustrations and silence make people want to believe.
I cannot fathom how it would not have been a Safari update fix by now if it was a problem solely with Safari. Similarly, I cannot fathom how it would not be an OSX update fix by now if it was an issue with El Capitan/Yosemite. This is why I haven't ruled out there's a hardware issue at fault. Apple have previously stayed silent on hardware issues until there was unsurmountable evidence of a hardware issue affecting a large customer base. That's usually the point that the fault was acknowledged publicly. See my post in this thread on the recent Mac Pro announcement as the latest example. Typically there's been complete silence in acknowledging the issue before this point.
I don't think it's worth making all sorts of wild accusations out of frustration from this. It's only a theory and one I'm only basing on educated guesses, not facts. There is also the valid point that this is only affecting a single action in a single browser, which makes a hardware issue harder to believe. So I'm not convinced either way yet.
If anybody working on HTML5 implementation on the open-source side of Firefox and Chromium would chime in that would be incredibly useful. Or those familiar with the deeper levels of WebKit. But that's wishful thinking.