I published my first scientific papers in the '70s, so I probably represent an elderly and conservative view on this. However, I would not use an idiomatic phrase like "pinch of salt" in a scientific paper - though I'd certainly use it in an appropriate context, in a conference presentation. I expect to see formal language in a published paper. Depending on your exact opinion of the results in question, alternatives to your phrase might be:
"these results must be viewed with caution because..."
"The significance of the results is questionable..."
or in an extreme case
"the results reported are inconsistent with well-established spectral data reported by several researchers [17-23], suggesting that the experimental procedure was flawed"
Of course, whatever wording you use, you must justify it!
I recall seeing a paper some years ago, pointing out a succession of inconsistencies in data presented in an earlier publication (by a different author), and ending with words to the effect "results reported in XXX should be trusted after, but not before, independent verification"