In the P&F era with test tube experiments and very small amounts of energy gain, I can see where measurement error was a big issue. I don't think a junior high science class could mess up the results on the scale of Rossi.
Nobody is talking about a mess-up. Mostly the assumption here is - reasonably skilled intentional fraud on the side of Rossi, and insufficient skills (or ability to modify the test) to detect it on the side of the observers.
Note that you have to also consider not only the skills of the observers, but the amount of authority to change anything that they have in these demonstrations. In most cases the amount of authority is zero. They are mostly coming as "good friend" guests which can not make "too difficult" request in fear to be kicked out and lose good relations with the demonstrator, and lose their moment in the spot-light.
Let's say you get a greatest genius in the world, put him in the photograph close to the apparatus and than claim that he was present in the demonstration. Does it mean he has _verified_ the setup or _organized_ the test? Not necessarily. It all depends how much leverage he has with the "demonstrator".
Let's say Danish professors would say - "we need to verify absence of trickery
in the lead shielding and thermal insulation by replacing it with our own."
The answer would be "sorry, for security reasons we can not do it".
So what does their presence or absence, or their skill level change? Nothing.
They have no ability to make any difference!
The only test that would be valid is a test completely organized by external
observer using their own equipment and 100% decision rights about any
elements of the setup. In this case, I quite agree, the needed accuracy
is quite achievable by a 5th grade science project level of skill and 500$ worth
of measurement equipment.
Regards,
Yevgen