Some of the greatest research ever done was done in Universities with grant money with no thought of any commercial applications.
Yes, but most of it wasn't.
Obviously commercialisation and testing is a different issue, that is where capital becomes relevant.
Commercialization and testing are almost entirely irrelevant compared to the capital required to research and develop new technologies...particularly in biology/medicine.
As with most things, this shouldn't be a black or white "patents are good" or "patents are evil". The question is how long should patent protection last and when should patent protection start. Some people think forever, others think zero years. The answer is most likely somewhere between those two
I will agree that with the rate of technological change today, the current 20 year protection is ridiculous. Technologies are typically woefully outdated by the time patents expire. IMHO patents should last significantly less time than currently (say 5 years or so), and should require that the product be commercially produced within some reasonable amount of time after applying for the patent.
Knowledge should be it's own reward.
Unfortunately, you can't eat, drink, live in or wear knowledge. At some point monetary compensation is required. The question is how much and how is this compensation provided.