If it can't be evaluated independently it's not science. At all. The defining feature of science, as opposed to superstition and recieved "wisdom" is that you d

If it can't be evaluated independently it's not science. At all. The defining feature of science, as opposed to superstition and recieved "wisdom" is that you don't have to take it on trust. (If it's merely difficult to evaluate independently, it might be legit, but you should still be suspicious of it.) Publication bias is admittedly more a problem.