But, if I understand your use of process simulation (and I am not sure I do), we aren't talking about people treating the mechanics as such. We are talking about a basic desire for the game not to disrupt willing suspension of disbelief. This is very different from wanting detailed simulation of reality. What I am saying is the designers can aim for playability, but they will also need to keep believability in mind as they do so. This may mean something like healing surge has to be removed because it produces too many believability issues for a fair number of gamers. But it doesn't mean the next edition of D&D needs to be rolemaster.
What I understand process simulation to mean is to distinguish it from result simulation versus some form of narrative or dramatic metagaming construct. D&D has never been pure in its approach to its mechanics. So we are talking about pressure to push an impure mechanic more towards one thing or another, not absolutes here.
The forms as they might exist in D&D might be thought of as roughly like this, using the canonical example of the swashbuckler swinging on the chandileer:
Process simulation - Mr. Swash has excellent dexterity (or balance skill or "swinging from ropes" skills--details are somewhat optional). He makes some kind of check, swings to a spot that the angle of the rope indicates is possible, bumps the orc mercenary over the rail, and lands gracefully with rapier extended, and then attack the evil gnome warlock. Each thing that was done, you do in order, using an ability and/or roll. If the checks are made, the effort is largely successful. If not, it gets interrupted (falling off the rope, missing the gnome, etc.)
The expectation is that following the process will produce results consistent with the desired fiction. If Mr. Swash wants to swing on ropes a lot, he will need sufficient skill to make swinging on ropes a high percentage task.
Result simulation - There are perhaps only three critical spots in this action--the swing, the orc bump, and the gnome skewering. (For more gonzo action, less critical bits, for more realistic action, more.) If the rope has to be deftly slashed loose before the swing, this probably isn't a check. In the process version, it would be. The critical thing is the swing itself, and anything else is at most a modifier, but mostly color. If the rope is thick, there may be a "swashbuckling" check or a general Dex check that is abstractly determining did Mr. Swash get where he wanted, how he wanted? If he succeeds, fine. If not, most likely the DM handles narration of some kind of failure or it is implied (i.e. a simple miss of the gnome, on that check).
Any number of gaming constructs, by themselves or applied to either of the above -- Fearing a failed check, or after already failing one, Mr. Swash uses an action point to improve the roll. He is just that good, as a heroic swashbuckler. The DM and/or player may or may not narrate using the resource, or it may be assumed to roll into fate or other thing difficult to pin down. One example of a purer versions (there are several), is that the roll doesn't determine success at all, but rather the parameters or consequences of the action. If Mr. Swash says he swings to the balcony, he so swings. The check is to determine how slickly he does this, whether he impresses his lady fair, or any number of options presumably important to the player and/or the character--and probably explicitly stated by the player before the roll.
A lot of times, the action in D&D is simple but abstract enough that you can fudge these differences in your mind, and it doesn't matter. But as pemerton has pointed out before, the Basic D&D saving throw is a form of the gaming construct that falls mainly in that third camp. You can see the difference with the orc in the otherwise small empty room:
1. The fighter rushes in and swings his sword. Almost all process simulation, albeit abstract. He moves per movement rate. He swings, rolling based on his skills. If he misses, either he misses entirely or the orcs armor made the blow null. If he hits, something substantial was done.
2. On the hit, the damage roll occurs, which is subtly result oriented. It doesn't matter how well he hit or where. If he rolls fairly high on the d8+3, he may take the orc down. If he rolls low, he will injure the orc but not take him down. It is entirely up to a person to narrate this in the fiction. (Again, not pure here, very subtle shift, which becomes obscured as the monsters get enough hit points to avoid take down from single hits.)
3. Then the wizard, foolishly fireballs the room. Thanks to the confined space, all three need to make saving throws. The wizard, standing in the doorway, can plausibly try to dive around the corner. But the orc and fighter only have their armor, maybe a shield, and each other to block damage from a blast that entirely fills the room. There is no simulated process or result that is consistent with the orc being fried by this and the fighter walking away slightly scorched. Nevertheless, with the gaming construct of the saving throw, the fighter may get lucky and escape this mistake--or not. If you can narrate something that works for you, you can gloss over this. But it is still there. Try hard enough, and you'll eventually come up with a theoretical situation that blows your suspension of disbelief--the fighter is bound hand and foot, spread-eagled, naked, and the wizard fireballs from 6 inches away, directly into his face. (It was a wizard that rolled a 3 on Wisdom.)
