I don't necessarily disagree with you, but imo it definitely matters how the problem is worded and presented, or if it is otherwise made clear what type of work will be accepted. Silly examples: suppose someone disproves the Riemann hypothesis through a computer search which finds a zero-crossing counter-example. Their counterexample is not a 'wrong' answer, even though they have essentially just done a more complicated version of 1^3 = 1, 2^3 = 8, ... and merely providing whatever computer code they used would probably be accepted by anyone. Example in the other direction might be the original four-color-theorem proof, which was partially by exhaustive computer aided search.