You made some very good points that were brought up elsewhere in reply to this post. So instead of repeating or copy/pasting them in response, I would suggest reading the top comment by smooth and the reply chain that follows and reply there if I'm off somewhere.
In regards to:
a) Dapps should be self-sustainable
I agree. And I see the DHF as a way to incubate worthy projects until they self-sustainable, and not a long term solution. But that's not all the DHF is accomplishing. Core dev, public infrastructure, and documentation are things that benefit us all, without a clear non-altruistic way to get funding.
b) A lot of projects are getting first class salaries for what they're doing from a community fund with very little to show as actual benefit for the ecosystem itself where I take the hbdfunding short term benefit 10x.
This is where it gets really tricky. Controlling variables is nearly impossible in these circumstances, so getting the data about impact of a particular project is usually circumstantial and tangential. Which is why allowing stakeholders to vote their confidence and letting it play out is the only reasonable solution. And the benefits of hbdstabilizer can be directly measured, but only assuming variables stay constant. If BTC tanks, then HIVE tanks, and it would produce a net negative while also losing all the things that could be created with the fund. What I'm getting at is don't put all your eggs in one basket.