I would think the only way to get a real good idea of how your die performed would be to roll it on a pice of felt a few thousand times and note the results. They might be slightly higher or lower but “never rolls a twenty” would probably be more important.
Here’s a test comparing a Chessex D20 to a GameScience (untumbled) D20. Unfortunately, N=1, so not a whole lot can be said. Still, the indication that we’re neither super far from decent dice, nor really very close to it, is quite interesting. Also, another thing to note is that they did not realize they’d have to remove a sprue from the GameScience D20, which gave it a bias.
i always used to roll all or many of the dice at the store , and selected those which ’ liked ’ me ~ nowadays , i just 3d filament print dice , and i can ( gently ) un-balance them as i choose , although , for most solitaire games , i prefer a biased electronic quasi-random number generator , carefully hand coded in 6502 assembly language
Used dice generators with my current tabletop group once, when we, through bad planning, were without physical dice. I have to say, while handy, they lack a certain something - and you don’t get that moment where everyone watches how a single D20 or D6 rolls, either!