Just to add to my following statement

All 4 sims yield the (pretty much) the same EV for IP (once OOP checks the flop)

I mean the there is a difference in the EV for using one betsize over another but that difference is so tiny (less than 1% of the pot) and I don't understand how it becomes a mistake to use one sizing over another.
The effort required to study which flop sizing is preferred on each flop by solver is huge and does not bring us any benefit so I am struggling to understand why a lot coaches now a days teach how to use the 'correct' flop cbet sizing.

I didn't read the article but when we are choosing the bet sizing , ranges are more important than board

For example at HU 100bb deep on AJ4r you can practically cbet range using 29%
But if you change ranges , for example input a stroner range for OOP solver will prefer a bigger sizing and a more polar strategy.

I think you are in the right way
You talk about villain responses and i am agree with you, population will make huge mistakes against small cbets

And most important imo, we will make less mistakes playing a simplified strategy

For example using a 75% size in a spot where is 0.5% better but is really complex will make us lose more EV than this with our own mistakes

However, if I am up against a good opponent who knows how to respond effectively vs each betsize, then it doesn't matter what sizing
I use ! Am I coming to the right conclusion here or I am missing something ?

This is true for the example but there are boards where EV diff is bigger for different bet sizings and the sizing you use matters

You need to start by checking ev of different strategies.

If you face a player who knows how to respond effectively vs each bet size you don't want to use a size with big ev lose

This is true for the example but there are boards where EV diff is bigger for different bet sizings and the sizing you use matters

This is exactly what my argument is, the EV will never be big enough for it to matter enough for us to investigate all boards and see which boards prefer bigger sizing.
Can you give example(s) where you think the EV difference might be big enough and I can run a sim to see. Happy to be proven wrong :)

probably you have to look in 3b spots because it will have much bigger pot on the flop than single-raise-pot. Therefore changes in EV between sizes can be significant

the situation in question for this discussion is cbetting SRP IP as the section of the article I referred to was : Mistake #1: C-betting small (33% pot) on Ace-high boards vs. the big blind

Hey pokerinlondon: what SPR did you utilize for your sims? The key reason to bet big on these flops is to get lots of money in with our more polarized range against Villain's more capped range. At 5 SPR, this is not a big consideration as we can still get it in with bigger sizings on turn and river. At 15 SPR, betting small on the flop no longer gets the money in, so the large flop sizing starts gaining a lot more EV.