The problem is that Steam is getting so filled with games (often of poor quality) that there simply isn't enough 'visibility' to go around.
The platform cannot handle the volume of games being added.
The question is: what is the correct number of new games, and how can that number be reached most efficiently?
I'd propose that the platform can digest about 70 new games each week, which is the number of slots on the 'new and trending' page. Currently, there are 205 launching each week (and growing), so there would need to be a reduction to 34% of the current level.
Curation is just not going to work with that volume. The overhead is too high.
Its clear instead that the $100 Steam Direct fee is too low. If we want to encourage professional-quality games (even from one person), this fee does not represent that level of quality.
I really think Valve should increase the Steam Direct fee to $500 - test it out, and see if that makes a meaningful impact by decreasing the total number of launches but increasing average quality. The $500 could still be recouped to the creator at a certain sales level.
To me the greater shame is the professional game launched to few sales, leading to the permanent exit of that creator from the industry, and perhaps their financial ruination, rather than a higher fee meaning a few less school kids and hobbyists uploading their weekend project.
All of this is poison to discuss online though, because many of the people filling the comments section on Steamworks and Reddit are the people who would be filtered out if the Steam Direct fee was increased.
Otherwise if the status quo continues, legitimate developers will face far higher advertising and PR costs (and time). If I could pay $500 or $5000 to launch on a much less cluttered and supportive Steam store, I would gladly pay that. This is the advantage that the consoles still have, and the success rates there are much higher because of the higher upfront costs and efforts keeping out low-effort content.