I'm going to go ahead and trust Harvard Medicine over a single study that has since gone largely uncited, and was published in "Nutrients" which carries an h-index of 115 (Nature, for example, carries an h-index of 1226) [0][1].
You should never judge an article by where it was published (some exceptions of course), only in the quality of the work. Impact factors are loosely a measurement of significance of the work, in terms of how broadly applicable they are to the fields of science. It is absolutely not a strict measure of the expected quality of the article.
Also, for what it's worth, most of the articles referenced in the Harvard Medicine review referring to dosages are from before 2010, so they may be out of date.
And why should I trust you? You've given a decree, "Thou shalt not judge a paper based on where it was published!" and then don't give any reasons why that would be the case.
Also, I'm not qualified to judge a paper by its work quality, I need to rely on the experts in the field for that, and a decent proxy for that judgement is where a paper gets published, so by looking at where a paper was published, I am judging a paper by its quality.
Sorry, I thought my reason was clear from my comment. I did not intend it as a decree you should follow blindly. I’ll expand on my explanation.
Impact factor is not a measurement of an individual article’s quality. It is a summary statistic, a rough measurement for how frequently that particular journal gets cited in other journals.
By judging an individual article by the impact factor of its journal, without reading the article and determing its quality yourself, you bias yourself to whatever those results might be one way or another with a statistic not intended to relay any information about the journal article itself.
You are right that it is a decent proxy where higher impact tends to have better journal quality (anecdotal, but I have observed this for sure) but I got annoyed when you claimed an article published in a low impact journal had no merit. That is simply not the case.
You can have the best, high quality scientific study ever about a highly specific thing, but because it is so specific you won’t have the broad impact that would get into one of the top tier journals. But with your viewpoint, that study would get thrown out the window.
Also, for what it's worth, most of the articles referenced in the Harvard Medicine review referring to dosages are from before 2010, so they may be out of date.