they can measure parts to know the size of features on one or several parts but there is no way to know what the tolerance is for each dimension
To know it? no you're right. To make a damn close guess? yes that is possible.
Take a company that uses the six sigma method. You can measure a sample of products and from that, take the standard deviation, find the sixth, and basically you can 'estimate backward' what their tolerances were likely to be. Instead of it being within six sigma of a design spec, you are going to within six sigma of what actually came out of the factory. Not 100% the same thing but in the real world, it's close enough. And before someone says 'the math doesn't come out exactly right' I will tell you, the math does not come out
exactly right, so you have to use your brain. Standard equipment for many engineers.
Considering any tolerances are going to have to basically be industry-standard anyway, unless there is a specifically patented manufacturing process that allows for better ones. And if so, you would know it because it would be patented.
For example, let's say Ford & GM both measure their cylinder bore to within +/- .000001 inch. Why would you assume Toyota can't also? Or Mercedes? The same machinery, tooling, processes are available to all of them. It is possible for them all to do it. Maybe some do not keep it that tight, but the fact that they could, you could design and specify the knockoff's tolerances to be slightly tighter anyway if it is reasonable.
Are they likely manufacturing knock off products with
tighter tolerances? no. Are they able to do the same? very likely. Are they actually doing it anywhere close? I have no idea, this is all hypothetical.
I like the composite wrenches linked earlier.
