[Mr 1936]

I agree with G8HQP Dave that there is no Fundamental problem in matching an antennna connected via a length of feeder. It will change the impedance seen at the far end of the feeder, but in principle this can still be matched to any other desired impedance given an appropriate network. There may be practical issues of course, such as the insertion loss of the feeder (which will be higher when terminated at both ends by something other than its characteristic impedance). The matching bandwidth will be narrower the longer the feeder is in wavelengths, which may or may not be important. Unwanted feeder radiation/pickup is caused by common mode currents, due to asymmetry in the system or the lack of an effective balun, again this is a practical shortcoming, not a fundamental one. A remote controlled tuning unit can be a heavy item, and it's not always possible to place it right at an antenna feedpoint, so some length of feeder in between may be unavoidable.

Antenna matching is important for transmitting, because normally the highest possible efficiency is desired. For receiving on long and medium waves, matching is often far less important, and often even with tens of dB of mismatch loss, the external noise delivered to the receiver will still comfortably exceed its internal noise. Matching only really becomes important if your receive antenna is particularly electrically small and/or your receiver is particulary insensitive. I admit that there is some satisfaction in seeing the S meter peak up, but at the end of the day it's the signal to noise ratio which determines how good reception is.