[silverbax88]

This is a valid thought, but I would piggyback on your comments and add that there are also engineers who put so much thought into their code - usually with the goal of eventual design expansion or reuse - that their code becomes over-engineered to the point that it, too, causes loss of overhead.

[chii]

There could be two explanations - one is that the engineer who "foresaw" the future requirement just got lucky. The other possible explanation is that they really had a black art in deciding what's needed, and what's not. And black art takes time, and experience, as well as some level of intuition (some might say its heuristic - but heuristic is just shortcuts for lots of gathered experience).

[com2kid]

Bingo.

I have done a good deal of studying of UX, I come from a background in testing, and I am now a developer.

Systems I design take all three disciplines into account. A junior UX team asks for something wonky in V1 of their spec, I code it to spec, and then make allowances for what V2 is likely to actually do. Typically a few extra fields in a struct, or designing a state machine to be a bit more flexible than strictly necessary. Maybe I break some logic out of the main class into a secondary independent one that can be passed around.