For our year 2000 retiree, withdrawing 4% from a 90% SPY, 10% cash portfolio would have been gut wrenching.

Indeed. (it would be unsportsmanlike to mention that the overall price/sales ratio of the S&P 500 is currently around 20% higher than it was in March 2000 when the S&P 500 peaked...)

It would also have been very tricky to implement.
You have to do some pretty serious financial analysis to decide when to use the cash versus when to sell more SPY to top up the cash to 10% again. (It's not THAT critical an issue if you get it wrong, but still, you have to figure out what to do). It also doesn't address the critical point of how you calculate how much you can spend in a given year. Yet this strategy was apparently aimed specifically at people who don't have those skills.

Unless a retiree is so rich they are sure you can live from the dividends alone, Mr Buffett's 90/10 it isn't really an answer to the problem of how to fund one's retirement from a portfolio sensibly.

Broadly speaking, there are only three approaches for funding a retirement from a portfolio.

(1) Pick a scheme that tries to run down the capital so you enjoy it, but not so fast that you run out of money if you live a long time. This simply doesn't work...longevity and markets are both very hard to predict, so there is always an unacceptable risk of going broke.
(2) Pick a scheme that spends no more than the increase in [probably smoothed] real value of the portfolio. You can never go broke, but you never get to spend the capital.
(3) Cut off the longevity risk with some pooled solution like an annuity or tontine. In a large pool of participants, not everyone is going to live an unusually long time, so each person gets a whole lot of risk reduction for a small fraction of the cost of funding it themselves as individuals. This might be immediate and with all the money, or a two-part as I have mentioned: spend most of the money in a straight-ish line for some years, then annuitize the rest at some advanced age to handle all years after that.

I eliminate approach #1 from consideration immediately. The dog food risk is too great.

There is a hidden side benefit to option #3. Every year that you live into great old age, you feel like a genius because the decision has worked out so well for you. The only circumstance that it turns out not to have been a good financial decision is the instance that you're dead, so you never regret the decision. So you spend your entire remaining lifespan chuckling at the chumps who took the other side of the deal.

Jim