Let make the assumption that you're getting the last 10% of improvement for ten times the cost and the mid-level cable cost is $250. It would make far more sense to have 10 cables that are capable of 9 out of 10 compared to a single 10 out of 10 while the rest are only 1 or 2 out of ten.
Having said all of this, I think the last 20-50% is mostly placebo or at least not much beyond different rather than better.
Having said all of this, I think the last 20-50% is mostly placebo or at least not much beyond different rather than better.