Impossible

Cell SPUs had explicit user managed cache and host DMA, and Intel ISPC and Unity Burst compiler/ECS (variants of C and C#) are good examples of programming environments that try to more explicitly encourage parallelism and/or optimal cache utilization.

Platform studies book series in general cover a lot of this https://mitpress.mit.edu/books/series/platform-studies.

If your goal is to get a job at FAANG or another big company early in your career you absolutely need a degree. For other jobs it will make things harder but it's not really a requirement. If you still want to work at FAANG 2 or 3 years of industry experience is worth more than a degree.

Game industry working conditions are generally significantly better than VFX and can be better than design firms looking for similar skillsets.

Edit: As stated in the reply Epic is now out also.

As some of my lawyer contacts pointed out, a lot of cases get settled out of court, and that data is much harder to get.

Anecdotally I've heard of and worked for companies that have been sued due to bias or perceived bias in the interview process. I have not heard of anyone suing or getting sued for feedback on technical interview performance, but in most cases there would be no basis for a candidate to sue.

That said I think the feedback suggestions are really off base unless the person writing the feedback personally knows the candidate. All of the suggestions are in the vein of "you clearly don't know X, you should learn it", which is fine if a candidate is actually completely unfamiliar with a topic. But the more common case, especially with basic CS stuff like Big O notation as used in the example article, is the candidate is familiar and got it wrong due to nerves, being rusty or interview time pressure. The same goes with the suggestion to recommend or buy a book for the candidate. Unless you are completely sure the candidate does not own the book already this is a terrible idea. Imagine how demoralizing buying, or even recommending, the candidate a book that they already owned and studied pre-interview could be.

I'm pro no feedback, but if the candidate absolutely insists I think feedback should be direct and based on the candidates interview performance, not perceived judgement of their skillset or knowledge (even if this is what most employers are trying to get from interviews). For example...

"You reached a working brute-force solution but did not solve the problem optimally. Here is a link on leetcode or hackerrank and a geeks for geeks solution to help you practice the problem."

"You did not come up with complete solutions to 2 out of 5 questions, even if given more time you were on the right track. Leetcode more and come back in 6 months"

But no one is that direct with interview feedback because the same set of leetcode questions that everyone asks are "company trade secrets". The feedback suggestions I gave are also not particularly tactful, but they are more useful given the current state of most tech interview process. I've learned a lot more about becoming a good interviewee from Blind and Hackernews than from any employer feedback. Richard Geldreich also has good insight into how the process works for senior candidates, and some of the political games that happen during the interview process that make any feedback even more of a joke.

For enterprise use cases I think VR will continue to grow in training, but teleconferencing and remote work (true killer business apps) won't land for a while. Whoever wins this market (my guess is either Facebook or Microsoft) will make a ton of money though. By land I mean, there are and will continue to be many applications that attempt to support remote work, but there won't be a useful one that gains enterprise adoption and is better than traditional video conferencing for a while.