I think a lot of the problem is the adhesive under the stones doesn't actually cover the entire surface. Every Chosera/Superstone which I've removed from their base has had, at most, maybe 50% of the interface covered in adhesive: the adhesive is applied along the long axis of the stone, toward the center; and has some significant thickness, perhaps 1/32". So there's a void along the edges of the undersides of the stones, where water will enter, remain, and absorb into the stone unless some little effort is made to remove it. This is compounded by the small lip molded around the top of the base, which catches water.
Furthermore, from what I've seen, Naniwa's production process always places the stones off-center prior to gluing; which means that although one side is right against the lip, the other side has a wider opening. This wider-open side happens to be the one above pictured.
Now consider how the stone dries if water pools under it -- it will have time to thoroughly soak in, and dry very slowly -- while the top, most exposed to air, has water drawn down and away from it by gravity: the top will be totally dry as the bottom will still be quite saturated. Or so it seems to me, anyway.
For what it's worth, I've been using Choseras with bases (400 to 5000 + baseless 8k Snow White) for around 18 months now, systematically drying them upside down, and have seen virtually not a single sign of even minor crazing -- the single exception being a tiny spider web once on my 3K.