I wonder if that's part of the story here too. Maybe it’s not just random dice rolls. Maybe it's that some businesses (the Mag 7 be damned), once they gain a foothold — through brand strength, ecosystems, scale advantages, whatever it is — actually increase their odds of pulling ahead over time.

More likely not, I'd say.
It is true that the US economy has a few insanely profitable giants (and a couple of less profitable giants), but a large part of that recent leadership effect is a one-time-in-history thing and another large part of it is those stocks simply getting more expensive. For example, look at the changes in market multiple of Apple and Microsoft in the last decade.

Consider:
There are extremely compelling reasons to believe that the largest firms are on average at any given time much more likely to be overvalued than a randomly selected firm is.

There is also surprisingly compelling empirical evidence. For example, in the 18 years 1997 to 20014 inclusive (i.e., ending about ten years ago), an equally weighted portfolio of the largest 5 stocks in the S&P underperformed an equally weighted portfolio of the other 495 by a rather startling 5.76%/year. Ponder that number for a moment. Could you underperform the market by that much even if you tried?

My contention is that the evidence for the recent anomaly continuing (the anomaly of the few biggest being the best bets) is pretty weak, and a prudent person of business would not wager on it continuing. While realizing that it might.

The normal thing is for leadership by size to rotate over time.

Jim