Hi people. Over the past few years I have crawled Brickshelf because I was worried that something like this would happen. I have uploaded the data to S3 and it is currently being served here: https://brickshelf-archive-7c8bea29901f.herokuapp.com/
The data in total comes out to about 2 TB.
Please note the following:
Search is primitive and only supports searching by folder ID or member name. Search is case sensitive.
I'm willing to host the site for a while, as it isn't too expensive, but may ask for donations. If it isn't needed because Brickshelf stays around, I'll take it down.
If I have time, I'll add some better browsing / searching functionality.