If the majority of research is tainted by conflicts, that doesn't somehow make it any less tainted by conflicts.
It's trivially easy to design a study to produce the results you want it to, just by selecting things like accounting methods at the outset.
For example, one method of accounting is to allocate costs based on usage. So if you have a car and you drive it 90% for Uber and 10% for your own usage, allocate 90% of all costs to Uber. That makes Uber look really bad.
But if you would have bought the car either way and the question is the incremental cost of driving for Uber, that accounting method is the sunk cost fallacy. It allocates large fractions of a bunch of fixed costs to the incremental use, like the original purchase price of the car, even though they're sunk and can't be avoided either way.
And the money actually does taint the study, because the funders know the parameters of the study ahead of time and only fund the ones that will produce the results they want, which makes the results tainted by selection bias.