My process is similar in that I test seating while 1st-fireforming of brass and at mid loads.
While seating can be tested for pretty much anytime, if you don't find it early then much of your efforts to that point might be invalid. That is, if you merely choose seating without a real basis and did a bunch of incremental load development with that abstract seating, you might end with a good load -while missing a better load. If you then decide to test seating, you might as well decide to redo all powder testing, as there is no way you'll know it's right without retesting.
IMO a reason reloaders generalize that seating matters less, or that the seating they happened to choose was best, is that seating adjustments from tune will immediately APPEAR to wreck a tune. This clouding their judgment about seating affects.
If they would stay disciplined in testing with an objective perspective, they would learn more.
But it's hard to be objective when right off the bat seating adjustments change what they've been working on... So they get scared of the really ugly groups and quit with the testing.
If a random choice is to be made, it should not be in seating.
Seating can represent the very largest affect to final results, both good and bad.
Powder pulls you in/out of tune, but rarely wrecks performance like mis-seating can.
This why I put powder charge as the random setting, while dialing in seating, then go to powder as a fine tune.
I like ladders here. If you're shooting ladders with random/chosen seating, well, good luck with that. But ladders can be impossible to evaluate while shot from a mis-seated load.