with the exception of Walmart, not a single top-10 company is still in the top 10, 35 years later.
...
But isn't this a factor IN FAVOR of cap weighting as opposed to equal weighting? If you cap weight, as the companies go out of favor, and as their market cap drops, the percentage they comprise of the index also drops.

But the value of your shareholding drops with it, while you are holding a really big allocation to it, making you poorer. So no, it's not an argument in favour of cap weight.

Rather than thinking about the trajectory of individual firms over time, think about the average return of your portfolio in an average year. That is a weighted function of the average return among companies in your portfolio. On average, a random company in a random year goes up in value a good amount. (more than the S&P 500 does, as a side note). The amount of return this makes you on average is almost completely unrelated to the size of the firm, so there is no good reason to prefer big over small or vice versa.

The one well known exception, and a very strong result, is the few at the very top in size that have a preponderance of overvalued firms, for very simple arithmetic reasons: any big company that gets hugely overvalued will
necessarily be found there. Their long run average returns are TERRIBLE. This may or may not be a good reason to skip them entirely, but it seems clear to me that it is plainly a sufficient reason not to want to overweight them. Do they do well for a few years sometimes? Sure, as does anything else. Is this a cycle, or a law of nature? It's a cycle.

Now, if you know one or more of those specific firms and have well reasoned conclusion that one or more of them will do particularly well, go ahead and overweight those. Or, if you believe that the fundamental economics of stock markets have permanently changed, and that competition and cycles of valuation don't really matter any more (for which a case could be made), then you my wish to invest in them. But neither of those is a discussion that has relevance to an index investor who wishes to get a decent return with minimum risk and minimum knowledge.

Jim