That is an interesting statement and it raises an important topic. You can verify your entropy gets included but what you can not verify with a computer is the amount of entropy actually present…
You can estimate the amount of entropy _you_ generate by accounting for all possible outcomes of some physical process you performed (dice rolls, entropia seeds, coin flips, etc) and in a perfect world your estimate would be exactly accurate. But no coin if perfectly 50-50 and no dice roll is perfectly random…but close enough, usually…probably…and if it weren’t you’d probably be able to tell…right? How? Well, if your coin came up heads 50 times in a row you’d probably get a new coin. But to rigorously quantify this?
Can you verify a computer is generating the amount of entropy you expect? This gets a little complicated and beyond the scope of my current understanding, but as far as I can reason the best you can do is output a large number of entropy bits from the computer and run statistical tests … but what tests and what sorts of statistical probabilistic / guarantees one can get and at what time cost I do not know.
