Repeatable might be an issue, that's true. But when one can claim the same numbers for pre-columbian americas, some african tribes and medival south-east india it would be much more convincing that it a constant about humanity.

IIrc Ray Dalio claims empires last 250 years. Why not 150 or 350? Is this roughly true for Quing China? Is it roughly true for for Cahokia? Or is it only for Dutch Empire, British Empire and Colonial Spain? A sample size of <10 is not a good indicator period - let alone if pre-modern age is even applicable at all.

I would be much more convinced if there was at least a large statistical analysis which he could have done with Wikipedia and Excel.