MangaDex used to have user activity based recommendations without issue on the old UX, which was the reason I and many other people started using the website at all.
It was incredibly easy to source new stuff to read by getting an objective metric of what people are reading in pure user volume, and data quality really didn't matter because it was pure clicks. I would be astonished if the current host can't track click volume at least for logged in accounts. If for some reason the analytics is busted irreparably and you can't pull pageviews without an act of congress (how does ad validation even work in that case?) adding a widget so people can click "banger chapter" and add it to a rolling tally for a week would be a sneeze and a half worth of time to implement in php. It's not rocket science, this is one of those things that was used as a practice example for comp sci classes throughout the 90s.
You can see an archived version of the old UX
here. As demonstrated in the sidebar "Top chapters" has a ton of stuff that was blowing up at the time, many of which remain popular today (or eventually got purged). There were transparent numbers of views that let you know how big something was getting in real terms, and you could sort views as far out as weekly if you wanted a slower pace of rotation. Current recommendations on prod are just what the same few ecchiheads staff are enjoying lately, a zany oneshot or two, plus things getting an anime next season maybe.
The UX has regressed in almost every way since the swap but most of those ways are much harder to fix, this one is the easiest slam dunk ever and it remains a relic of the past after 6 years. The sidebar already scrolls to fit all the junk on it, put the pageview widget in there and wham you get community curated manga picks again with twenty minutes of work.
Even this thread has been going for over 3 years and nothing.
All I can do is bump, I guess.