Vote sellers will adjust which will make buying votes cheaper

If buying votes is cheaper that means vote sellers are getting a lower return. Someone else is getting a higher return. Congratulations, you have now helped shift the balance in the reward pool away from vote selling! This unavoidable.

Don't let the perfect be the enemy of the good. Yes, there will still be vote selling, but it will be less blind, and less profitable, with the balance going to more meaningful rewards, content and value discovery, and to non-milkers. Take the win.

But its not both as in the reasons are equally present. Its not fifty fifty, its something like 95/5..

To be perfectly accurate here, it is still both even if not 50/50. But in fact there is no way to know the weighting without changing one of the variables and running the test. If we reduce the cost of downvotes and downvotes hardly increase because of the factors you describe, then we will know you are correct. If they do (meaningfully) increase we will find you are are incorrect.

I don't pretend to know the answer, but I'm very interested in finding out, rather than guessing or pretending to know. We can then go from there.

Economic incentives have a very limited effect

In fact, economic incentives have a great effect, or we would not see precisely the sorts of behaviors we see now, which are nearly optimal given the current incentives (and which changed a lot when we changed the incentives before). Change the incentives and one of two things happen: The behaviors will no longer be optimal under the new incentives and the milkers will lose out (and therefore, inevitably, non-milkers will gain), or behavior will change and your claim will turn out to be incorrect. Again, I look forward to actually seeing the results instead of guessing or claiming to know everything, but either way it is still an improvement.