Understanding, and inspiration. They had to create under serious constraints in compute, memory, and storage, and understanding how and why they did can lead to ideas about how to optimize software on modern machines.
It's also critical for understanding how and why the engineering choices were made when documenting the evolution of processing. Instruction sets, processor design, programming languages, computer culture, corporate trends, all of those things have roots in design decisions, and the software preserved on tapes like this are a sort of DNA.
The effort needed to incorporate the information is dropping, with AI you can run analysis and grab important principles and so on, and whatever principles govern optimization and performance under constraints will be useful on a permanent basis.