No, they are not. He never claims they are made to Casino quality. He claims they are more accurate than the competition for gaming dice...and he's right, they are.
As this independent analysis indicates. They concluded, "Crystal Caste translucent dice are clearly the least consistent dice, Chessex and other styles of Crystal Caste appear to be more consistent than Crystal Caste translucent but less consistent than GameScience and Koplow dice, with Koplow opaque dice representing a “middle of the road” option." So, no scam. If you want more consistent dice, you should buy GameScience or Koplow. If you don't much care, then Chessex or non-translucent Crystal Caste is fine. And translucent Crystal Caste is the worst.
Plus, every single time you get a gamescience die, you have to sand off the extra elements that are always on them. It's like they were all molded on a large rack and then broken off, so there's extra bits of plastic on them, that you have to sand off. The end user sanding off plastic, doesn't sound like it would do a very good job at ensuring the dice get a scientific outcome.
And this I find hilarious.
ALL dice are made this way. It's just that competitors are sticking the dice in essentially a rock tumbler...to round ALL the edges, including the sprue. You are much more accurate in carefully removing the sprue, than any dice company that is bulk rolling the entire die, sprue and all, in a polishing device. Because you're just removing the sprue - not all those other edges in an essentially random method.
If Chessex uses a process of: molding, then cutting, then polishing, then inking the numbers.
Gamescience only does two of those, then somehow tries to convince you its better.
First, Gamescience does offer inked dice, it's just up to you which you buy. Second, yeah, "polishing" means "dumping them all in essentially a rock polisher" to make all the edges rounded. That is not a "feature" for many people, it's a "bug". It means some edges are more flat than others. It means stacking those dice on top of each other will get you a more random height, not as fixed a height. And that's provable.