I suspect that part of it was because there was a brief time when there appeared to be (at least, in certain important segments of the non-enterprise market) a consolidation around a fairly narrow set of frameworks, tools, and approaches in web development, and that time roughly corresponded with a point in time when web development itself was really taking off and drawing lots of people in -- this roughly corresponds with the height of Rails ascendancy for the server side.
As a result, there's a lot of people who started web development at a time when it may have seemed, if not simpler, less diverse in terms of strongly-supported options -- and trending toward greater consolidation, rather than greater diversity. For those people, the current state of web development could legitimately be a surprising array of more-complicated-than-they-would-have-expected decisions, with lots of opportunities for analysis paralysis.
As a result, there's a lot of people who started web development at a time when it may have seemed, if not simpler, less diverse in terms of strongly-supported options -- and trending toward greater consolidation, rather than greater diversity. For those people, the current state of web development could legitimately be a surprising array of more-complicated-than-they-would-have-expected decisions, with lots of opportunities for analysis paralysis.