I think we will get around to extracting it but there are a lot of roadblocks to doing anything with it. A lot of it is IP is a complete mess so even if a company or institution did want to digitize it actually making it available is a complete nightmare, it's ok on '23 to '78 works because it's at least a fixed length but on any works printed after that the period is variable so you have to research when every author died. [0]

[0] Not even sure how that works for things like journals or pieces with multiple authors like scientific papers! I imagine for collected works each piece would be under a separate copyright duration?