Also the FACTS are that rebalancing is primarily cosmetic and has no real effect.

Though that does sound like a good general conclusion, I do suspect it depends on the rebalancing strategy in question.
I am open to the notion that some good ones might exist, notably if they are based on valuation levels and expected returns rather than targets for percentage of portfolio.

Consider this one:

From time to time, check the relative valuation of the two asset classes you're considering. Not frequently.
If one is unusually cheap and one is unusually expensive today, and the mismatch direction has moved away from the situation the last time you rebalanced, then rebalance today.
Otherwise, sit on your hands.

As a worked example, check out the long term data on the ratio Berkshire Hathaway's fixed income holdings as a percentage of book or of non-subsidiary investments.
Fixed income was over 42% of investments in 2000, and about 3% in 2021.
This isn't because management finally realized that fixed income is a bad choice, it's because the expected returns on offer tanked so the prudent allocation changed.

It goes without saying that the baseline allocation should be a lot more in the asset with the higher long run expected return, and zero to whichever of them may currently have a negative real return expectation.

Jim