I keep noticing something which looks odd to me - I am sure it is
correct because this appears to be standard practice, but I would
appreciate it if someone could shed some light on the reasons...

Take a peek at http://web.telia.com/~u85920178/tx/500w-txt.htm (first
example I found). Specifically look at the 5 position switch on the
output matching, in position 5, the bulk of the inductor is shorted
out. This just looks plain wrong to me.

Hi Mark,
If you take an inductor, make a tapping and pass a hugh RF current through
the smaller end, then a voltage will be induced in the other side. If the
tapping ratio was (say) 10:1 rati,o and 300 volts was developed across the
1, then the bigger end would have 3000 volts at the other end = arcing and
"sparkalating", smoke and switch breakdown :-(

Regards