I agree that layer 2 offers a way to handle this situation and once layer 2 becomes more established, layer 1 may look drastically different, with the reward pool much reduced as people leave it to move money into layer 2 communities.
For me, the issue is both content reach/discovery and also rewards. POB is specifically designed and sold to join together post reach with rewards. Content creators in Web 2 (let alone web 3) are used to being able to monetise content in order to power their content creation efforts and this is part of why the reward pool is largely directed to content creators, by design. I shouldn't need to point this out, but it seems that some people like to warp the situation to suggest that anyone who suggests creator rewards are important is some kind of scrounging excrement. Youtube is what it is, in terms of size, in no small part due to it's focus on paying content creators and Steem was designed to capitalise on this idea too.
In order for POB to function, it requires that content creators are both visible and rewarded - based on community sentiment. In the early days of Steem it was agreed that whales would not drop heavy upvotes (or downvotes) that much (or ever) in order to give POB a chance to reflect genuine community consensus, but that has fallen away now - so that those involved have decided that pure stake is more useful to Hive as an indicator of community sentiment than is the actual voting patterns of large numbers of people. This has caused the age old dynamic of money vs. people to surface here and people are polarised as to which is the 'right' approach. Typically, those with the most money choose 'money' and everyone else chooses 'people'.
Despite how it may look to those who don't know me personally, I am actually in the camp of 'doing what is best to grow Hive' rather than 'doing what is going to get me the most short term rewards'. As someone who is specialised in both systems engineering and business/marketing, plus specifically in social networks, I am very aware that public perception is fundamental to the success of any social network and that actually people are more interested in social dynamics than they are money earning potential, overall - when it comes to adoption of social networks. However, since Hive absolutely merges these two things, the money aspect here automatically knocks into the social dynamics, meaning that the way that money influences the social dynamics on Hive is a fundamental factor in it's growth potential.
All of this is leading to the relatively simple point that we can't successfully disconnect post discovery/reach from the rewards and economics of the system. Not only are they totally connected at the blockchain level, but people are tired of being exploited on Web 2.0 and they are not generally going to accept being told to use Hive UIs that simply ignore downvotes in order to accomodate whales who profit heavily from their content generation behind the scenes, but who are actually downvoting them constantly for their own reasons. People do regularly scoff at this idea and mock the idea of participating in such a system.
I am just doing what I can to advocate for system design and community cohesion that makes the best use of the reward pool from a growth and marketing perspective. Having been on the receiving end of heavy ideological downvoting now for months, I can appreciate what so many people here have been telling me for years in private - that the downvoting can be a huge problem when used in ways that are not optimal for growth (a polite way of saying it). I see reducing the amount of free downvotes as being a simple, low cost way to explore solutions to this (a quick fix) that may actually work. This has never been adjusted in all the time it has been active and I think it is high time.