Ah, come on guys, I think you are overcomplicating this.
From what I understand it's basically a test rig for flexible cable moving back and forth and the goal is to capture any glitches in the cable as it moves.

Using a PIC with IOC will probably work fine as long as the code for it is written properly. The only thing it will NOT catch is a second (and third and forth and....) fault condition on the very same line that initially caused the interrupt. But then again the absolute fault count per line wasn't that critical and I think that the interrupt latency etc will provide the "perfect" amount debounce (even though the specification said no debounce) for this to work perfectly fine. But again, make sure to read up on how to handle the IOC events in order to not miss events on lines any of the lines that didn't actually cause the interrupt currently being serviced.

Using a port expander, which I also suggested in my first reply, will work but in this case I don't see the benefits, only drawbacks, considering Barry did find a PIC with the needed number of IOC pins.

/Henrik.