I just quick-looked at this, and don't see any particular problems. Here are some comments:
The document is dated 2018 - it is likely not some ancient work, although the principles and components have been around for quite some time.
I don't see any problem at all in using the classic old-school parts - I do it all the time, instead of buying new stuff - to help use up some of the millions (not an exaggeration) of components I have accumulated.
The part about impedance matching seems to just be lamenting that sometimes people don't pay attention to it. But they will have to, when it causes problems. I think in this instance it was about the mistake of connecting the nearly zero Z of the buffer emitters directly to a 50 or 75 ohm line. The "6 dB" thing is encountered all the time - if you have a low Z thing, say nearly zero, the simplest broadband match is a 50 or 75 R in series to the line, giving up half the available voltage. For a short run, you can do tricks to make it OK, but for a long line, it's best to just do it right and take the loss in exchange for better signal integrity. It depends on the frequency content. So, I don't think this was about different cable types, but simply the need for proper matching.
Ed
