You're absolutely right, it does have its limitations, and they show in the issues you raised with it in a way that doesn't actually apply to the nature of the analogy as it criticises the original situation under examination. The nature of indexing is such that it's treated by practitioners as if it were an "objectively correct" (and obvious, no question about where the volume is flowing, the numbers are right there.) answer to the correct investment allocation, so it isn't actually so much the case that there are people just copying random other picks because they like their haircut or it's a popularity contest or whatever (although I suppose you can still make the argument that you're making some kind of decision in an index based on
which index fund you're allocating your investment to, but even there you could take a no favourites approach by indexing index funds.)
The core point I was trying to make though is that the market is efficient in net across a great variety of actors all making their own judgement with regards to how much a given thing is worth because it's assigning so many eyeballs and independent evaluations to the pricing question for a given item.
When you swap that out and replace it with a system in which n% of the trade volume is just copying the rest of the trade volume, that reason applies inversely proportional to the volume that such indexing takes place, until a market with just one guy calling all the shots and everyone else indexing him may as well just be centrally planned, and thus afflicted with all the heinous cancers thereof.
Your point about the deleterious effects of the above also simultaneously lowering the bar on the challenge of beating the average though is something I hadn't thought of, and is completely correct. I guess that is how indexing would collapse when you got to some state where the market was mostly just ignorantly following the tiny minority of organic judgement being exercised, and that tiny minority turns out to be inadequate. The market then recovers by stock picking once again being conducted by specialists, and as you say, this would be a situation wherein stock picking was massively undervalued, as opposed to the (maybe present?) situation where it was massively overvalued.
Thanks for the book reference, I'll definitely check it out.