I will not argue that you seem to favor such theses.
I think your only example that was NOT reinforcing "card is not resolved as though it occurred during the M/H phase" was Doubled Vigilance. It seems like the important aspect of Doubled Vigilance is that it's passive triggers off of a
declaration ("chooses to enter"). That is, once something has been declared it has been declared---chains resolve what was declared not whether declarations occur. A good point/purpose for the rule wording.
I think the other historic example was Lure of Power being revealed in response to an influence attempt being successful, which has a passive with similar characteristics.
But the Lure of Power brings up its own aspect to consider. The on-guard reveal is allowed to occur
during chain resolution. Even if you say the on-guard is
literally resolved during the current chain while just its effects are
treated as though they occurred earlier, that still doesn't allow anyone to respond if the reveal occurs during a chain resolution. Konrad's suggestion would fix this, although without giving much of a reason and perhaps leaving ambiguity about handling multiple on-guard cards.
Konrad, would you consider something like:
"A revealed on-guard card is the first declared action in a nested chain of effects, and when resolved its effects are implemented retroactively as though the card was declared prior to the chain(s) of effects during which it was revealed."
In retrospect, this seems to me to be likely the intent of the original rulemakers, just insufficiently formalized.