I don’t think you did the experiment you think you did. Your parallax error values for half of the scopes are way off. The max parallax error equation that you cited is accurate. It is based on simple geometry and the accuracy of the equation is insensitive to assumptions. The primary assumption is that light travels in a straight line. Over these distances, I think we can accept that assumption as valid. Readers should see the following link, in which the equation for max parallax error is very clearly derived:
Derivation of Riflescope Parallax Equation
For a scope adjusted to zero parallax at 100 yds, which should be the case for most if not all of the scopes you tested, the max parallax error at 300 yds would be equal to the objective diameter (1.4-2.2 inches). I don't need an equation to calculate that. The geometry is very simple. Yet you reported a parallax error of 7-8 inches at 300 yds for half the scopes.
The only way I could explain your results would be if the parallax free range for half of the scopes was a lot shorter - about 30-35 yds! I doubt that is the case, although I guess it is possible. It seems more likely to me that your measurements are off. If your measurements are correct, then you should send half of those scopes back to the manufacturer and have the parallax free range properly set to 100 yds or some other value that you prefer.
Scope power doesn’t matter.
All of the modern 6-20ish scopes that I can think of have either AO or parallax adjustment that will eliminate the parallax error at virtually any target range if properly adjusted. Maybe there is an odd one or two out there that have no AO or parallax adjustment. If so, they are rare and not worth worrying about.