There is a way of 'doing science' which has long stood the test of time. It doesn't matter how many people support a theory, or how eminent they may be. Nullius in verba - take nobody's word as truth.
Nor is experimental 'proof' to be trusted, since the search for proof predisposes one to seek verification and validation of one's own biased view. Experimental disproof is the key. It takes only one well-conducted experiment to prove that a long held belief is false.
Robert Boyle explained this idea of experimental disproof in terms of coins, which in his day were of gold or silver.
" For I am wont to judge of opinions as of coins : I consider much less in any one that I am to receive, whose inscription it bears, than what metal 'tis made of. Tis indifferent enough to me whether 'twas stamped many years or ages since, or came but yesterday from the mint. Nor do I regard how many or how few hands it has passed through, provided I know by the touchstone whether or no it be genuine, and does or does not deserve to have been current. For if, on due proof, it appears to be good, its having been long, and by many, received for such will not tempt me to refuse it. But if I find it counterfeit, neither the prince's image nor superscription, nor the multitude of hands it has passed through, will engage me to receive it. And one disfavouring trial, well made, will much more discredit it with me than all these spurious things I have named can recommend it.".
Prefatory remarks in A Free Inquiry into the Vulgarly receiv'd Notion of Nature, made in an Essay adress'd to a Friend: In English and Latin. (1686)