People always trot out the percentage as if it's going to definitively prove me wrong. When you're moving from 2.5 to 3.5, yes it's trivial. It's 1 freaking point. The 36% is nothing but ONE POINT. It's like when news folks try to scare you into listening or reading an article by telling you that if you drink a beer your chance of butt cancer goes up by 100%, but then you realize that your chances went went from .000001 to .00001 or something.
Which would be a 900% increase (x10)...
... sorry. But that had to be corrected.
A 100% increase would be x2, so
.000001 to .000002
Other than that I mostly agree.
Optimizers value to hit bonuses more than damage bonuses in general. This does not take into account what you showed with your example.
In a white room, a +1 to hit bonus increase your damage most (by 100%), if your chances to hit are 1 in 20. So you go from 5% to hit to 10%.
In the white room, that seems fine, in actual play you should not bother to hit with that kind of chance.
Also a +1 to hit bonus only shows in 1 in 20 rolls.
And if you are a level 20 fighter with continual advantage and action surge, where that increase (5% to 10%) might actually show, you probably did something wrong to begin with.
On the other hand, damage bonuses are stronges when your chances to hit are already good. And it is not a 1 in 20 chance where you get the jackpot, but you have a steady income which often is better, even if the average increase is lower (standard deviation is also lower, which is good most of the time).
The only time I'd clearly prefer to hit bonuses is when you can apply a very debilitating status effect like stunning strike. Just attack as often as you can and just try to get it in once. And if you did it is an instant win.