Concept of minimal exceptionless justification

I was often wondering how to decide if you want to play a candidate move or think a bit longer about the position, and I was also thinking about different ways of justifying a move and how to do that efficiently.

I was fascinated when seeing how in a few videos where strong players discussed positions and arrived at the same conclusion via different justifications, as well as in a few vote chess games. It leaves me wondering if there is any value in the concept of finding the optimal justification for a move and therefore excluding other moves quickly which didn't meet the expectations and finding the correct move.

Assuming there are justifications that make it clear one move is correct, what should be the methods to find them. The other demand is that the justification is exceptionless, meaning that you can't arrive at a wrong move with it in different positions. If it is too spongy for that, more conditions have to be added that demonstrate why in this case, and maybe not in other cases, one move is the best and the other isn't. If you use non-exceptionless justifications, you are basically just taking a guess, even if it might be an elaborate one.

In tactical lines and with forced play, it can be the easiest justification to give the critical variations

For tactical positions and forced play, this can sometimes just consist of giving the critical variations, maybe shortcutting it with giving a plan and include why the opponent can't stop it.

For positional play, and even intuitive tactical play, there is a lot more reasoning and application of positional rules required, plus certain arguments why you can apply a certain positional rule in that specific situation.

The method of exclusion also can be a justification, but generally one needs to have a good reason to play the moves one played before when not having any reason but the method of exclusion for the followup.

Some examples (although they are not good ones, because I didn't really prepare this post, and also I'm a weak player, so I can't teach anything about this but hope to learn with it myself):

White:

Kf2 because every other move loses the rook, and there is no compensation in any form for losing the rook, because we only have a knight and don't have a knight fork or a tactical shot.

Black:

Here it obviously is more difficult, for me it would go along the lines of:

Nd5, because it opens the d8-a5 diagonal for the queen, and with that indirectly prepares an eventual flight from the king. It can also help as a defender against the h5 ideas.

Well, that didn't sound as obvious and reasoned as well, and I'm pretty sure there are examples where all of the stated would be true and Nd5 wasn't a good move. So it probably either needs a clean positional argument (I don't know one) or a lot of lines, maybe there could be a compromise with a bit of calculation and some positional argument.

Another example to think about:

Black to play:

I tried to copy this idea, originally from mathematical argumentation, where you can also use proven theorems and a bit of calculation to construct a new proof, but you need to have the right idea and argument, out of an enormous set of possible arguments, and there are often elegant proofs that are much quicker than the "brute force" calculation method. Although the equivalent of the proven theorems in chess are not really proven but rather general principles, most of this translates into chess. There is always some degree of uncertainty left because of these general principles, but I tried to limit that by demanding a justification for the possible application of that general principle in this position.

If you have any questions about the concept (because of my terrible long sentences and bad choice of words, probably), I'll try to make it more vivid.

Please share your thoughts, and if you have a bit more of a clue about chess, feel free to post better examples with good move justifications. Maybe you also have some "elegant solutions" with the right thought process to solve a complex position quickly ;)