TVL is not the goal; TVL is a necessary evil. Every dollar locked in a (especially immature) protocol instead of being locked in a simple and secure wallet is to some extent a security risk.

Sometimes we have to take that risk to get valuable benefits, but locking for the sake of locking should be minimized.

then how do we pay for the research and development costs? How do we identify the potential valuable benefits down the road to justify the up-front work (that's getting increasingly hard to find and pay for)?