I am in the process of curating and eventually uploading some of our datasets to a public platform called DANDI archive. They support BIDS, so the transition will be simple, but we need to de-duplicate the data from our internal database. In #351, I see that we can have a mirror to public datasets, maybe this would be the way to go. Here is a list of what I plan to upload:
also, this one (but it's hosted on github here)