I still don't think that aspect alone would be as challenging financially or technically as it was back then.
I am one of the few that had some exact numbers regarding the financial costs of both series 8 and geek week. Unless I've missed something the cost hasn't changed (taking inflation into account). The major expenditure arises from the slow filming process, which burns huge amounts of studio time. When you have a crew of 100+ and a full day's work (due to the use of chromakey) only nets you 20 minutes of footage, this is the cost that quickly adds up.
That also answers the point about technical difficulty. If you want some real numbers on that, geek week acquired about 40 minutes of footage (less if you cut the long introduction) and took two looooong days to film. My team's quest (20 years prior) was about 35 minutes (from memory) and filmed over 2.5 much shorter days. In other words, the process today is no quicker.
(Theoretically, it's always been possible to do it quicker by simply assuming that the void will align with the virtual environment. That's what we did at the convention. But you may have noticed that the alignment/perspective was never quite perfect. Therefore, to do it properly, you have to manually align the rooms to the virtual environments, and tinker. This takes time.)
What the pace of technology has done is put basic chromakey into the hands of amateurs. As you say, YouTubers do this endlessly. But it's not comparable because it's not the chromakey itself that costs the money, but more accurately it's the consequences of using chromakey that costs the money.