Indeed. The point being you're supposed to implement the latest and/or proven and/or best techniques but we're reduced to sloppy work due to business requiremen

Indeed. The point being you're supposed to implement the latest and/or proven and/or best techniques but we're reduced to sloppy work due to business requirements almost always. Your point about encryption is misplaced. Do not imagine impossible stupidity - even ancient Romans sometimes encrypted tax and military information. The difference is between brewing your own (common in 90s, crackable even then in a day of math or so) and using actual research. As for collaboration, it is overrated. Best work generally is done by a few actual geniuses in the specialty. (I can name a bunch in a few fields I'm interested in, like acoustics and audio algorithms.) They do need support staff of course. Usually the real ones are tenured professors with long development of a singular technology over the years. As for penicillin, it was one quite genius chemist (Pfizer; Erhardt was a comfectioneer so support) who developed means of mass producing it. (Among many other things later.) They worked for a company which got major government funding during wartime, which makes it almost like a tenure.