BT252 is a 20Msps ADC, and four of them connected to an FPGA an I/O accessible via a standard connector could be a basis for quite some nice experiments. The PCB is probably relatively easy to reverse engineer, I assume the rest is some TTL buffering or similar.
The problem with these things is that it takes a lot of time to write software for it, and after you've invested all that time you have one piece of hardware. If that dies somehow you're out of luck.