Well, life happened and I was unable to respond to this topic as quickly as I wanted to. Hopefully some follow-up is still useful, especially since PKM still has not received any notable priority 
(aside from Readwise integration maybe?).
I of course love that this scenario is free, however I personally need (and Notion growth was incredibly supported by): free collaborators. Not full accounts, just collaborators, if that makes sense. In other words they have their own free, solo workspace, I have mine, and I don’t manage or pay for their account, but we can collaborate in each others’ workspaces in some way. Maybe that’s just Comments or something, ideally more. Look at how Notion handled/handles it. And for that I would pay e.g. $5/mo. But if you can do that for free, all the better. 
It’s a pretty poor task manager with notifications the way they are. For me personally that’s a big blocker. Until recently the desire to create “dashboards” that mix content, like different views, free text, etc. was also a missing piece. Now that we have Columns and embeddable Views in regular Rich Text we are much closer to this. 
Further improvements to the block editor would be welcome though, e.g. more of A few should-be-easy Rich Text features to add from Remirror - HR, mixed lists, collapse lists. Columns was one of those, so it’s great to see that came around.
But the other thing that is probably my biggest issue is not unique to Personal knowledge management (and I honestly can’t understand why it hasn’t yet received more priority): better tools and default settings and whatnot to deal with way too many effing databases more smoothly in daily interactions. The prime example is free text # link search, where I believe it would be extremely helpful to be able to set per-Space or at least global Workspace defaults for types to only show in linking (by default), with a “show all” as an option, but the default filter would show, well, a very small set most likely (I have 25 total Databases, I commonly link to less than 5). A related issue, which I should report as a bug, is that the # link pops up a list of Recents, but as soon as you start typing it’s sorted by… god knows what, but it’s not recency, and it should be IMO (or at least optionally).
The overload of databases became especially egregious when I added Google Calendar integration into my workspace, which is now completely “polluted” with one-off calendar events and reminders, forcing me to either remove integration, or change how I use Google Calendar. I want it integrated so I can e.g. see all my Tasks and my Appointments on a single calendar, which could be extremely helpful, but it’s quite uncommon for me to want to link to those events in free text. The same and perhaps worse happened when I added the more recent Readwise sync. It’s even more infrequent that I want to link to something like a highlight in free text (for my personal use case; I can imagine for others it could be quite common!).
Let’s keep in mind, of course, that linking to specified databases through relations is much more common and more focused for specific relationships. As a result I almost feel like link-in-free-text database search should have to be turned on for each DB rather than turned off (this is a little too extreme, but I do almost feel this way because fully 80% of my DBs I basically never link to this way, I just use normal Relationships, which already have good filtering, etc. built-in). More discussion of this in Ability to hide some databases from search/Entity # Link by default.
Speaking of all that, the final thing I’ll bring up, again not unique to PKM, but perhaps exacerbated by those use cases, is that Search really needs some work. I’ve made numerous feature requests about this so I won’t repeat myself, but I’ll mention that I think serious PKM users will tend to have a lot of Databases. Indeed one of the primary draws for Fibery vs. Obsidian and other Markdown-based tools is the ability to more clearly delineate different “types” of data. And any look at how people use Notion and the many templates available will show you that for just one area of a person’s life they’ll often have several databases. Now add all those areas of life into one PKM, realizing the promise of Fibery, and you suddenly have this massive problem of actually finding what you’re looking for, both in regular search and (as mentioned above) in link search. It’s a serious impediment that I think may be almost worse in PKM use cases, even though of course a full-fledged business use of Fibery could include 10s of DBs too. Just recognize that PKM uses at least as many and you can see how the search problem is critical to this use case. I have seen some improvements here since the start of this thread, so that’s nice, but I do think a lot more work is needed (perhaps “semantic search” will answer much of this, assuming it’s something I can just use by default and don’t have to pay for API access over time).
I was planning to use Fibery’s built-in Template sharing. Since I have many connected Databases in many Spaces it would be a full workspace share, though you can choose to only import parts. Unfortunately I am further from being comfortable doing this than I hoped I’d be by this point (see “life happens” above 
). But I do believe and hope that there would be pretty good value to some folks once I’m able to do it and your interest is certainly a motivator. I will see if I can make a plan for a phased to polish up my existing systems and maybe plan a phased sharing of what I’m doing. I’ll note that I bet people like @Sarah_Arminta and @YvetteLans probably have similarly or even more expansive PKM setups they might be able to share some or