[kopos]

Took me rushing through all those years of insane coding in the last company I worked for. By the time I had quit, a fully functional game engine had undergone 3 rewrites with 2 more underway and with one of them not seeing the day of light. All for the same reasons you elucidated and some more.

While it was not 'idiots' that was the reason, the later developers always had just enough more data on the assumptions to be made for the current revision. But then again, not enough to foresee the need for a fresh rewrite when things changed again in the future. Sometimes, the rewrite was essential, sometimes it was just on a subjective whim to build a silver bullet once and for all!