If I remember right, the first time when entering with Oculus, there are over-lays for each Touch controller.
Don't get me wrong. I
love those VR controller call-outs you see in various games. That is a very elegant system. However, I believe that is probably provided by the device sdk. He just has to stick in the labeling he wants.
That is somewhat different from what is being suggested here.
Those call-outs are always there and there is a fixed number of buttons that need to be labeled. And the labels are static.
In this case, the callout come and go. They need to be dynamic. Which means they need to be scriptable. They need events to inform them when to show and when to hide.
The following is not what you'd have in a beginer training, but the idea is the same:
https://bbs.hitechcreations.com/smf/index.php/topic,397806.msg5273544.html#msg5273544 Your training mission would need to script those call-outs to show up at a certain time in the mission and you probably don't want to blast them all up at once, so there would need to be script/event structure to say "show a call-out on control X at t+1:35" Then OnCLick event to tell the system to close that call-out and show the next step in sequence when the button is clicked. Etc, etc.
So you'd need a scripting and event system for every button and listbox item on the UI and some identifier for every control widget on the UI so you can refer to them in your script and invoke a call-out, as opposed to a video where you have graphically annotated stuff with existing tools like:
I know there is some stuff like this in the mission system but I only remember seeing controller button and axis events etc, but I don't see anything involving the clipboard UI wigets.
So, if that kind of capability already exists, could easily be created, or he was planning on doing it anyway, then rock on Garth!
If not, it sounds like it could be a lot of work to implement a new framework.
But as someone else said earlier, ANY in-line training you can put in front of a new player is going to be a huge benefit. I'd say sooner the better. If a perfect system would take months and months to implement, then I'd suggest simpler approaches you could do quicker first. See if it starts to pay dividend, THEN refine.
Nothing is chiseled in stone. I'm sure training would evolve over time. Later, simpler versions could be replaced with slicker approaches.
Perfect should not be made the enemy of good, and sometimes a less perfect solution that arrives on time is superior to the perfect solution that arrives too late.
$0.02.