Seeing that the connections were quite simple, and I already made a RJ10 breakout, it would be a simple matter of wiring up Vcc, Gnd and RxD and seeing what happened. It was all TTL 5V anyway, simple and easy to deal with … and this time, for the full experience, it’s a video. Enjoy.
So it turns out, Kaypro’s engineers were right – their keyboard cable mustn’t have been regular phone cable. The reason the keyboard didn’t work was the resistance (they wrote impedance, while technically not incorrect, did make me think more of signal/AC problems where it is more likely that there was too much resistance and voltage drop) in the thin and fragile phone cable.
Now that Kaypro keyboard cables aren’t available – how can one get around this? Well, one way is to cut off the phone cable plugs near the end (so there’s only little stubs of phone cable, thus only adding a minor amount of resistance) and then soldering in a thicker short cable in-between the two ends (say CAT5 cable, the two conductors in a pair paralleled to reduce resistance). I haven’t quite got the time and motivation to do this, but knowing that the keyboard still works – that gives me hope that I might do it one day.