I've tried it, it's tough to do. Too loose and it dries up in days. Too tight and it doesn't transmit water. I gave up on it after a few experiments, especially since it doesn't offer many (if any) advantage over cotton.
This is what I ran into initially.
What I've discovered is that I can drill a cork that is the inner diameter of the test tube. For example, I'm using test tubes with a 16 mm inner dimension. If I cut out a cylinder of PVA that is about 16 or 17 mm in diameter, it fits when both completely dry and when completely saturated.
I had to get a number of diamond-coated hole saw drill bits to get precisely-sized cylinders of PVA, though. They're super cheap but annoying to have to get more specialized tools. Also, the sponge has to be completely dry before you drill out your PVA cylinder.
After I got the right sized PVA cylinder, I had to try various heights for the cylinder to determine which would allow moist sponge on the dry side. If the cylinder is too tall, no water makes it through to the dry side of the PVA.
The advantage I'm hoping for is reduced molding. We'll see--like you said, probably won't make a difference.
Edited by prettycode, September 28 2015 - 10:45 AM.