I have not, but I assume that the market will vary beyond a simple mathematical function like that. If the median were accurate enough, you could pretty easily establish a mean reverting target based on the HLC/3 & HL/2 of, say, the first hour. Anything that straightforward could eventually be arbed, changing the market's character.
The reason I prefer VP to the pivot is two-fold. On one hand, it's a tool derived from auction market theory (mfbreakout quoting Steidlmayer saying that you want supply to define price discovery, not price), which conceptually makes sense to me. On the other, Fisher said in at least two places off the top of my head that his calculation was simplifying what Steidlmayer was already doing. In the book his colorful illustration of price moving through the pivot is a sword piercing a fat man's belly -- the smoother it goes through, the more confirmation that price will continue. That bell curve image is of a volume value area. In one of the videos, he talks about the pivot as approximating Market Profile without having to pay for it. So I figure, just use the real deal rather than the abstraction.
Just eyeballing it has shown how the pivots and the VPs don't match up at times, so that's enough for me. But they do often appear close. Since I'm not going to take the time to test any of these variables, I'd rather try to get conceptually close to the ideas behind the theory as possible.