I would say your problem is the framing :) You're assuming an outcome because you know antibiotics work. What is the incentive to spend $10m on research for a particular drug? It's not "because we know it ends up in a cure for X", because we haven't done the research yet. It's "because we think this will reach enough people to be useful and we'll commit a lot to this thing vs lots of other options."
Let’s also imagine the other side of the equation. Can you not imagine any penalty or cost other than bankruptcy? Let’s say you are forced to allocate $1 million per year to research. is the only cost function you can imagine based on the risk of default?