These are great points and certainly are aspects of Findings we would like to improve. I have a few things to say about this topic, which hopefully will be of interest to you, and to prospective research groups.
First, I would be remiss if I did not mention the discounts we have for purchase of group licenses: Group Discount — 20% or more 
Now, let me just clarify the kind of workflow I have in mind and that works today for sharing and collaboration with Findings. When a user is done with an entry, they export it as an archive to a shared folder to which you have access (Dropbox, or anything else, for instance a network drive, or any internal system set up by IT for file sharing). You can then manually import new archives as they come in. Findings will not create duplicate entries, and will instead merge the latest changes when if a new archive is put in. The main drawback is that this approach is manual, but it also offers the most flexibility.
One potential evolution from this setup is to make everything automatic: any update is pushed automatically to the shared folder by the version of Findings running on their machine (with potential settings as to when this is done), and on your end, Findings detects new or updated content and load it into your machine. This setup can even have an electronic signature system that alerts you when an experiment you already signed is changed by the user, avoiding the problem of silent changes to older data. In other words, signed experiments would be locked, and there would be no way for anyone to alter the data, since only you would have access to that version on your computer.
At this stage, the above is of course just a plan, and I can’t promise a date at which it would be possible, but I sure hope to get to it in the future. I’ll update this discussion when this happens, but in the meantime, feel free to comment and share ideas etc.