I came across a link to Always the Same Warning Signs (science.org) while browsing Hacker News.
It goes over some of the warning signs around revolutionary (and lucrative) commercial science breakthroughs, like in pharmacology. Specifically, what if it seems too good to be true.
The piece gives three warning signs which ring true in an engineering setting:
- Only one person can get this great stuff to work
- Legitimate questions are met with stonewalling
- Important data are missing or kept secret.