The cooler does look very good, indeed. It was done by someone who knows what he does
Attaching peltier elements to a rare card, without to know if it helps anything when it comes to OC, this my big problem. This should have been first tested on a normal 4500/5500 to see if the VSA100 does have more clock resources at all.
I think that the watercooling solution wasn't at it's limit and a simple voltage boost from 2,7 to 2,9v could have shown if the cores can be clocked higher. We know that the VSA100 are also able to work with 3,3v for a short time!
Why not test the watercooling v6k to it's limits first? Why start to mod another v6k without knowing that there is a benefit and do all this effort and waste the money (an the card...)? This is something I simple don't understand.
All those cooling modifications could've been done with just one card. You see the point?
Adding Plastik 70 here is just the top of the whole story, by the way. With just a little bit research in the web one could've found way better solutions than using this kind of stuff. This is my opinion.
Having enough v6k does not excuse overhasty actions
Ok, from my side everything is said.