As others have pointed out, there is no central repository for front end analytics, which means that there is no way to accurately determine the number of views a post achieves network wide that come from outside of Hive. Sites could voluntarily forward their statistics, but confirming their accuracy would not be straightforward. Additionally, do we really want to trust Google with having a hand in the distribution of the Hive rewards pool? (Google Analytics is the industry standard currently).
There are open source analytics systems that could be used, but they could also be abused.
I think monitoring the onboarding and activity of new users is a more reliable approach to tracking marketing success of different users and apps - but that's also not straightforward and the more rewards can be gained by it, the more effort will go into gaming it.
As others have also kind of pointed out - since quality is entirely subjective, we end up with the content being rewarded that the community likes. Since it is all stake weighted, those with more stake shape the narrative and visibility of posts on the network overall. Despite claims like 'no-one looks at the homepage' (totally untrue), the reality is that rewarding posts well does make them more visible in many ways.. and this is a genius idea.
There are several common problems I can see that result in missed growth/SEO/marketing optimisation opportunities for posts and lists on Hive:
What one person may personally like is not necessarily what the rest of the world likes. So truly quality curation that drives traffic will take into account current trends and the posts that drive traffic. Doing this requires having an eagle eye on SEO trends and social media stats from Web 2 sites. I am not aware of any curators that have even spoken about this on chain.
As you pointed out, posts that go viral on web 2 are often quite simplistic and/or they come from 'big channels' - meaning they are produced by content creators that have large followings and decent production budgets/skill. Neither of these categories are well respected on Hive. Simple posts are likely overlooked for being too 'lazy' or they are downvoted for being 'exploitative' if they get upvoted. The only channels I am aware of that post 'big channel' content on Hive are @tdvtv and @cast.garden - who either syndicate content on behalf of 'big channel' users or actually have those users using their site. These guys take the biggest downvotes currently - EVEN WHEN THEY SET THE POSTS TO BURN THE REWARDS! Just go and look at the history of Vigilante.TV and Cast Garden (or monitor the untrending report that tracks downvotes to see for yourself). So clearly, the bigger curators are missing big opportunities from a content optimisation perspective. It's true that there are counter arguments, such as that the content creators don't often engage the community here - but genuinely, having listened to the few of them that have commented on this (james corbett, max igan, pressfortruth etc.), they have all specifically said that they are put off from using Hive due to all the downvotes. Even after the situation is explained to them they still choose to stick to other networks as it is a matter of principle for them. Something to consider.
In short, there is no agreed upon marketing strategy and little appreciation of SEO or viral marketing among the curators from what I can tell. This is all reinforced by the lack of real world data to clarify what is really going on to the curators.
This seems to be an area where centralisation actually offers an answer, in that front ends which combine analysis of the real world data they collect from web 2, tied in to referring hive accounts, stand a chance of curating the right content to grow the network (and their own site). If they can also mitigate any downvotes they might attract in the process, then they are probably going to grow more than other Hive UIs.. maybe that's the best we can do for now.