1. Spellbook, from the vault and similar products
I see some Premium features around collecting. Maybe they make adding CSVs not wanted.
I'm more of a "use free stuff and write some own tools that suit me" rather than "pay for solution that adds to already expensive hobby" guy.
But definitely love what deckbox offers for free accounts and don't want to screw around their monetization model.
2. Dropbox vs Git
a) collaboration
Anyone can easily clone repo, change file contents/add files and propose changes, either fixes or new CSVs
with PullRequest they present their "version", you can see all changes highlighted, and either accept them as main version or not
and even if you accept changes that were incorrect there are tools to revert unwanted changes
(its easier or harder based on how they were merged and what changed after them)
I don't think you can set up dropbox like that
Downside is you need to learn some basic git commands and github UI.
And migration - if you have all files in some kind of directory you just commit them to git hub repo to some data folder.
If they are scattered you'll need some script to find, download and store them on some reasonable paths.
If you are doing anything around tech i really recommend learning git anyway as it is a really powerful tool for sets of text data like csv or code.
b) versioning
you have history of all changes saved, who and when changed them,
so if any file gets corrupted by mistake you can easily revert to file version from a few months away,
of if you see e.g. one person was sloppy when contributing you can review their changes later to double check them
also if you need to e.g. migrate as some CSV column changed name or format, e.g set name column "Rivals of Ixalan" was changed to use 3 letter code "RIX"
you can put changes in all files as a single "commit" so they are all marked as connected
and not only latest version is available, but all files before and after change can be downloaded if needed
git does the job of comparing all versions and storing only changes, so repo size doesn't grow too fast, but all versions are available
you can have multiple variants leaving in their own parallel universes called "branches", mix and match changes made on them
you can even try applying some changes made to old versions to a different version
(this requires both versions to be somewhat similar or it will not work reasonably, but magical merge algorithm works surprisingly well in most cases)
for simple CSV this is a bit of an overkill, more useful for working with code,
but even for CSV it allows more control, easier cooperation and experimentation
simple scenarios are pretty straightforward - you jus commit to change locally and push to change on server
but when you start adding multiple branches and merging different versions it can get complicated
and tools support all those different scenarios so they may be tricky to use some times
but i see you already host on git hub so you know at least the basics
c) organization
all CSVs are in one place and one can download repo or part of it to hdd if they want to
and if they want to run some scripts e.g. to import deck to Arena they can easier than in dropbox model where they have to use site or script crawlers
3. UI
It's a good practice to separate data and presentation.
So that you can create new presentation without altering data and add data without worrying about presentation breaking.
You can have multiple presentation options if you have separated clearly. e.g. One page with a table and another with expanding lists visualizing same data file.
We have a lot of tables in my work project so i suggested going with them as They allow filtering by multiple factors, searching etc out of the box.
But default setup looks worse than i remembers so definitely needs more work.
Not sure how good you feel about scripting. I'd put most layout work in Javascript and provide data to display in JSON, CSV etc.
This does require some learning to handle. And you have to unify structure, limit to a few styles rather than write custom layouts for all cases.
But then you can easily experiment with different changes and process data with scripts as layout and data are not coupled.
This is more of a mind set to learn - like in the script above,
thx for adding more info,
but when storing data like Category, Type, Note its much better to use different fields and glue them together in JS
than to concatenate all to Text field and have no option to change layout later or present list of all Categories with filter by category
4. General cooperation
I can prepare my version of https://meldon44.github.io/mtg-precons/
Hosting it temporarily on my account https://NamespaceV.github.io/
And if you like it I could PR to your https://github.com/meldon44/mtg-precons repo
I have some time ot experiment now as im not working for some time an d project seems fun,
but wont be able to support project in the long run.
PS
github.io is great, i was looking for some lightweight free hosting and this is really nice
No server side scripts but other than that really great - mimics CD pipeline without any devops for static content
and JS + static files can do a lot if you like that stack
