If one had to pick between SPY and Berkshire for the next 10 or 20 years, and close your eyes, there is only one answer, SPY.

I, myself, own way more Berkshire than the SPY now, and each quarter I sell off a piece and reinvest it in the SPY. I sell off 3.2% per year of Berkshire. It’s not really denting the position, I may need to go to 4-5% per year.

Berkshire without Buffett will look very different, with untested leaders.

I think it will prosper, but who knows if it matches the index.

It is tricky.

Earnings Yield: Berkshire wins.
Berkshire earnings yield about 6 to 7%
SPY earnings yield about 4%
Interest rates are higher than the earnings yield. SPY is expensive!

Retained Earnings Test: Berkshire loses
As Rationalwalk points out, Berkshire has failed to meet the (revised) retained earnings test in 8 of the last 10 5-year periods.
It's on track to fail again this year.

Book value not a good measure anymore? I agree. Let's use the Buffett-approved 5-groves method:

Rolling 5-year periods comparing SPY price vs 5-groves IV:
-3.8% 2013-2017
3.6% 2014-2018
1.2% 2015-2019
-1.5% 2016-2020
-2.5% 2017-2021
3.2% 2018-2022
-3.8% 2019-2023

Berkshire has lost out to SPY in 3 of the last 6 periods and is on track to lose again this year.

On market price Berkshire has lost for the last 4 periods and is on track to lose again this year.