I have a firm belief that the original point-and-click adventure boom was an accident. A lot of previously popular but now niche genres follow a pattern: They applied some constraint to their interaction that lead to (relatively) high-fidelity graphics and/or storytelling on weak hardware and with small budgets. Once technology advanced to the point where said graphical fidelity could be achieved without those constraints, all but a few hardcore players left for other genres. Fighting games (where the machine could spend its entire resource budget on two characters and a largely static stage), rail shooters (where the fixed camera and scripted encounters allowed the designers to very easily know how many entities would be onscreen in any given frame, even to the point of being almost entirely pre-rendered), racing games (cars are easy to model, fixed tracks give the designers full control of entity visibility, sprite scaling allowed for 3D-ish effects before 3D acceleration existed), flight simulators and space games (the sky/space is easy to render, planes and spaceships are also easy to model compared to humans). None of these genres are as popular as they once were, probably due to the fact that even a go-anywhere open world game with hundreds of living entities on screen can run on modest hardware these days, so there’s no benefit from a graphical perspective of applying those restrictions.
The adventure game and the visual novel once occupied this position. By being a series of (potentially entirely pre-rendered) stills that could only be interacted with in a few predefined ways, you could make impressive-looking scenes even on 16-bit hardware. It would be a very long time before anything as impressive as Myst could be made to run in real time on consumer-grade hardware. I suspect a lot of people only bought these games because they looked pretty, and once that edge was gone, they realised that maybe they didn’t really like the genre in the first place.
I think that there's another factor to consider. The market for computer game was both much smaller in the 80s and early 90s but also tended to be much more educated. So, games that are essentially puzzles with stories would have much more appeal to that group of people.
I think also culture was much more ready to put "mindful" and engaged deep thought time on content. As lot of games due still end up burning lot of time, but I feel they often are more mindless or more in the moment gaming. Kinda tetris like deepflow. Our attention spans have shortened a lot and there is lot more distractions available.
That sounds plausible. Or at least it was part of it, because at least Myst would never have sold anywhere near as many units if it didn't look so impressive (in 1993).
Though there is also the simple fact that pure puzzle games can't compete with puzzle+action games (e.g. Zelda), even if the puzzles are somewhat reduced.
You can see a similar thing in movies: Highly successful Pixar movies are typically comedy+action, because mixing in action beats pure comedy in terms of entertainment.
The trend towards action is also present in other genres: Turn
-based strategy games got mostly replaced with real-time strategy games, and turn-based RPGs got replaced with action RPGs where fighting is done in real time.
The adventure game and the visual novel once occupied this position. By being a series of (potentially entirely pre-rendered) stills that could only be interacted with in a few predefined ways, you could make impressive-looking scenes even on 16-bit hardware. It would be a very long time before anything as impressive as Myst could be made to run in real time on consumer-grade hardware. I suspect a lot of people only bought these games because they looked pretty, and once that edge was gone, they realised that maybe they didn’t really like the genre in the first place.