So here’s a framing that I found useful, maybe someone else will too.
Given some problem area, let’s say is the importance of the problem, defined as the total value we gain from solving the whole thing, and write for the proportion of the problem solved depending on the total resources invested (this is the graph in the post).
Now let’s say is the amount of resources that are currently being used to combat the problem. We want to estimate the current marginal value of additional resources, which is given by .
The ITN framework splits the second factor into tractability and neglectedness. If we write for resources normalized by the current investment , then
The factors on the right-hand side represent tractability and neglectedness . So we’ve recovered the familiar = marginal value of additional resources.
But this feels like a kinda clumsy way to do it―it’s not clear what we gain from introducing . Instead, we should just try to estimate directly (this is the main argument I think OP is making).
There’s a variant of attitude (1) which I think is worth pointing out:
b) Progress studies is good and we should put resources into it, because it is a good way to reduce X-risk on the margin.
Some arguments for (1b):
Progress studies helps us understand how tech progress is made, which is useful for predicting X-risk.
The more wealthy and stable we are as a civilization, the less likely we are to end up in arms-race type dynamics.
Some technologies help us deal with X-risk (e.g. mRNA for pandemic risks, or intelligence augmentation for all risks). This argument only works if PS accelerates the ‘good’ types of progress more than the ‘bad’ ones, which seems possible.