I'm not sure how much there is to learn from "the DM probably won't let the bad guy crit multiple times in a row".
Risk management. If multiple crit runs are off the table, one can more confidently work with typical damage numbers when considering the risk vs reward of adventuresome actions in combat. E.g., "if I rush to save the Wizard, the evil Princess and her thugs might hit me, but I think I have the HP to risk it...and they shouldn't crit so much that it would kill me." (Note that risk of death is still present in this example; it just won't be "eating three crit OAs in a row" levels or whatever.)
Unless the player hates player side metagame mechanics.
All of my suggestions have been for DMs only. I have not recommended any metagame mechanics intended for direct player use, and in general I am trying to avoid metagaming at all (hence the emphasis on diegetic solutions).
Does the safeguard give you at least one of the negatives you listed above (how do you learn things from an ad hoc safe guard that may or may not happen again) .
What kind of safeguards were you picturing?
I gave an example earlier in this thread (IIRC, anyway) from my own Dungeon World campaign. See, I tend to be a big old softie when it comes to DMing, and I was worried my players might feel Il was throwing fights at them that weren't meant to be more than speedbumps. So I set up a combat meant to be slightly tough but not too hard...and the party demolished it no sweat. At that point, I decided I needed to make a fight that would (at least I hoped!) be clearly too hard. As part of that, I built in diegetic "safeguards" that could be exploited later to make the fight less dangerous. Specifically, these were bound shadows (a favored trick of the Zil al-Ghurab, the Raven-Shadow assassin cult), which were already known to have magic that could manipulate life force. I intentionally overloaded the fight with smaller "minion" creatures, led by one greater shadow. After the first two rounds, the party realized that they were in a bad position and needed to either break or achieve a major success, so they chose to nova strike the greater shadow (presumably intending to disrupt the lesser ones, they didn't explicitly say). In one round, they were able to pump out enough damage that the greater shadow WOULD have died right there, but that would only have partially mitigated the difficulty of the encounter, given the sheer number of other opponents.
Instead, I creatively exploited the life force magic of the shadows. The greater shadow, sensing defeat was near and deciding that it was better to flee so it could maintain its mission in the future rather than be destroyed here, chose to absorb the life force of (most of) the lesser shadows and then tried to book it, hoping to lose the party and then reconstitute the smaller shadows later. Instead, the party gave chase and destroyed the greater shadow. They had a very real brush with lethal danger. However, because I prepared with a possible countermeasure to mitigate the encounter difficulty, I was able to address the problem purely diegetically without needing any metagame mechanics nor fudging rolls or stats. (I can't actually fudge rolls as players make all rolls in Dungeon World, but I obviously could still fudge stats. I just don't do so.)
Going from the lowest tier to second highest does feel like a thing
And some people liked 3.5 and loathed 4... ::

:;
Oh I know, though I would prefer not to delve too deep into that. I merely gave it as an unequivocal example where I know for a fact that a player (that is, I) wanted something from the game but genuinely did not know that I wanted it. I had had no idea that what I really wanted was a rigorously balanced system that put every player on a level playing field and said "alright, gloves are off, do the best you can with the tools you have."
Sometimes, we get caught up on what we believe something to be, or we fail to realize what specific thing is actually valuable to us because we just don't have the experience to give it a name or a shape. That doesn't mean we should just roll over and tolerate it if someone says "play something else." We should, however, be open to the possibility that the tool we have become attached to is not effective for the purpose we intend.
But I think we three of us agree that the "go play other games" tangent is done, so I'm willing to leave it there if you are.
Unrelated, this was in my +Quote backlog and I figure it too is worth a reply.
I thought everyone would be more accepting of fudging because they TRUST the DM's JUDGEMENT. Afterall, they got that EMPOWERMENT for a reason?
Well, firstly, you should know that not everyone is particularly keen on all this "DM Empowerment" rhetoric. (As far as I'm concerned, DMs were already plenty empowered, and nothing actually changed about how much power they have or had.)
As to your core points, I don't think it's that weird to be upset about finding out someone is concealing things from you, and it certainly wouldn't make me less suspicious if they followed that up with "but you're supposed to trust me!" More importantly, I have seen too many examples of well-meaning but faulty DM judgment over the years, both personally and from testimonials online. No one is infallible, but fudging that requires such "trust"--that is, secretly modifying rolls or modifiers after they're already in play--by definition has no error-correction mechanism unless the secret gets out (and almost everyone agrees that getting "caught" doing this is a Bad Thing which should be avoided). So it both projects a (false) impression of infallibility, and denies the only possible means by which one can get feedback and error correction.
I'm cool with
cooperating with a DM--but cooperation, like respect, is a two-way street. Expecting unlimited trust is just as much of a problem for cooperation as expecting unlimited approval. Expecting trust to be backed up by communication and collaboration is not a bizarre paranoid reaction. It is not
weird to respond negatively when you find out that the DM has secretly manipulated results to be what she wished them to be,
even if she had the best intentions, and only did it wishing to help the players.