> I would imagine (but admittedly do not know) that a cosmetics patent holder would have little reason to refuse to license to some entity operating in a very different field.
That's not really the problem. The problem is that you have entrepreneurs at the early stages of looking into developing something who discover that it's patented. Just the overhead (paying attorneys) to negotiate a patent license can be thousands of dollars, more if you want to negotiate a good rate. Nobody wants to pay that at Stage 0 before you even do the legwork to determine if there is really a promising product there. But determining that can itself cost many thousands of dollars and nobody wants to do that before knowing they can secure a patent license at a reasonable rate, so you have a chicken and egg problem.
Meanwhile there are many promising alternative avenues of research that aren't already encumbered by patents, so people go there instead and come back when the patent is expired.
That's not really the problem. The problem is that you have entrepreneurs at the early stages of looking into developing something who discover that it's patented. Just the overhead (paying attorneys) to negotiate a patent license can be thousands of dollars, more if you want to negotiate a good rate. Nobody wants to pay that at Stage 0 before you even do the legwork to determine if there is really a promising product there. But determining that can itself cost many thousands of dollars and nobody wants to do that before knowing they can secure a patent license at a reasonable rate, so you have a chicken and egg problem.
Meanwhile there are many promising alternative avenues of research that aren't already encumbered by patents, so people go there instead and come back when the patent is expired.