You may have posted this elsewhere, but I'm curious if you have a model that projects the same (expected returns over 7 years), but for RSP instead of SPY.
On the the hunch that the 'magnificent 7' are currently skewing SPY's valuation metrics considerably, and the remaining 493 companies have considerably better prospects for something approximating normal market returns.

No, I haven't done it. Not much.
It would be much more predictable, but it's harder to pull together the data. I could do it, but it's a bit of a pain.

I agree that the concentration in markets these days is quite remarkable, so maybe I should.
And the high concentration in the biggies in the US is causing a high concentration in the US within the world.
From 2006 to 2012 inclusive, US equities represented 45% or less of the global index (MSCI All-World). Now it's about 62%.
The latter is partly valuations: in that old stretch non-US equities traded at an average valuation discount of around -10% to US stocks, and lately it's averaging about -30%. That's not the same as saying that non-US equities are "buy", but if one can figure out the explanation then there would certainly be implications for investment strategy.

The magnificent seven were not obviously more overvalued than most of the S&P 500 pretty recently because of of them were pretty reasonably valued relative to their huge earnings, but that's not longer the case. (emphasis on "not obviously"...maybe they're worth it, maybe not--different discussion)
As of about a week ago:
Cap-weighted average earnings yield of the big seven equates to a P/E of around 29.8 times current run rate net earnings. As usual, Amazon and Tesla are the outliers, both over 75.
Cap-weighted average of the rest comes to about 15.4.

Jim