The program itself requires two files to exist before it can operate. Then it asks for the input number in base 10000 as a set of digits separated by spaces or returns. This is pretty bizarre by today's standards.
It does succeed in factoring very tiny numbers, but it seems to have bugs which prevent it from working for large numbers.
The double large prime variant of the MPQS seems to allow numbers up to a pretty small bound. This strongly indicates that it is not competitive.
I'll see if I can get anywhere with the double large prime variant version (edit: I tried the second program and it fails with the same runtime error), but given the extremely poor state of the code (no comments, full of gotos, no indication of structure, no documentation, no test suite, buggy, very poor interface, etc.) I would say this code is not going to be particularly interesting to modern day researchers.
At the same time, it is a tremendously amazing accomplishment that one man essentially toiled away in secret and created a massive number theory library like this! When I google his name I find nothing. Was he associated with an institution? Did he publish any papers? I know he didn't use the web, so that explains why he doesn't have a web presence, but it is odd not to find him referred to anywhere. Or am I just looking in the wrong place?