Besides being a great testbed for database coding, I think such a project can only success in the community, if it comes with value added.
Just offering an addon downloader interface - like OH already is - will not attract a lot of people. But the OP already outlined a possible feature that many folks wanted: easy addon management. If you combine the downloader with a manager (i.e. easy installation, no config fiddling, proper collision solver, no missing references etc.), it has the potential to lift off.
Unfortunately, such a combination would mean to not only proxy downloads from OH, but to inspect them, analyze them, rearrange them, add meta-data and so forth. This is a much bigger coding challenge then writing the equivalent of an exchanged CSS of a given website.
I think the reason why there is "hostile" attitude towards the OP is that most know this aspects and assume that the OP is ignorant to them, thinking that it will just come together along the road. TBH, I tend to fall in this camp, too, from what I've read here. Without the "hostile", of course...
my :2cents: