I think it in large part comes down to how the performance assessment process manages PMs at an org like Google. When you're measured at work on your success in 'launching' things, and the evaluation metric comes down to your personal idea authorship and impact generated, there's a profound need to extinguish other paths. An incremental improvement is insufficient, you need to have launched / authored / designed, and provided personal outsized impact. Or as an engineer, played some leading role in making this vision reality. The organization as a whole is constantly on the hunt for these units of work, because the internal metrics encourages this. X launched and replaces Y is a good story to tell the promo committee. Iterating on someone else's design is career suicide.