Since we release all our data on figshare (so there are DOIs and a permanent archive), you can use their great API to download any of our releases.
All of the releases are under a collection, which you can see here:
https://figshare.com/collections/GRID/3812929
https://doi.org/10.6084/m9.figshare.c.3812929
The last part of this is the collection ID. Using the API we can view the data about the collection with the ID:
https://api.figshare.com/v2/collections/3812929
The contents of the collection can be accessed by adding '/articles':
https://api.figshare.com/v2/collections/3812929/articles
From the list of articles, you can find the most recent and then use the url there to get details, for example
https://api.figshare.com/v2/articles/5032286
Under "files" it'll contain a download link, for example:
https://ndownloader.figshare.com/files/8500607
For recent releases, this will be the zip file containing all the data. For our very first releases, it'll be either the CSV or JSON files.