It is a system with a feedback time of months-years. At these time horizons, you can’t do rapid iterations, and you have to be extremely risk-averse. The system is also indeterministic so not conducive to hacking.
My application for EB2-NIW was recently denied for totally inappropriate reasons. The process has a randomness component to it, what you can do is to maximize your chance of success but it’s still a chance.
That’s false. You can apply for EB visas from outside the US. After the approval of the I-140 (application for employment-based visa), instead of “adjustment” of your visa status in the US, you perform immigrant visa processing at your local consulate to get your green card.
You can still travel after filing the I-140 but not after filing the I-485 (adjustment) without advance parole.
I am an F-1 doing an EB2-NIW so I have some background around these topics.
1. The bar is very high. You must show that you have "risen to the top of your field" through very substantive evidence. There is a list of admissible evidence on the USCIS website. An attorney can discuss whether this is worth pursuing from your credentials.
2. The USCIS values citations but it's not everything. In my petition, I relied on reference letters from experts in the field who have cited my work from across the country. Again, an attorney can help with this.
3. You must not have immigration intent at the start of your F-1 period. However, you can change your mind afterwards and you can file an I-140 after 90 days, which is what USCIS considers a "long enough" time period. But in doing so, you have demonstrated your intent to stay and it's highly recommended that you do not leave the country until you get your GC.
Not attempting to claim malfeasance or anything but PNAS has a weird "contributed article" track (this article is one of them) in which the authors _can pick their own reviewers_ [1].
"Only 18% of direct submissions were published in 2013, whereas more than 98% of contributed papers were published, according to figures on the journal's website" [2].
Note that the authors compared blood erythritol level, not _intake_. Erythritol is also a metabolite, which is produced by the pentose phosphate pathway, and has been associated with metabolic dysfunction before.