[G8BBZ]

Picking up on a comment by Radio Wrangler, improvements were not limited to receivers. The improvements achieved in frequency setting accuracy and stability in transmitter exciters was also a major driver of change.
In the good old bad old days, it was not unknown for a transmitter to drift by anything up to 1 kHz after a frequency change, before finally settling somewhere near the nominal channel frequency as everything finished warming up. If the link was in traffic at the time, the transmitter would have to be carefully chased up and down the band until it settled. Crystal controlled transmitters were obviously better, but so much more limited in what they could be used for due to the finite number of crystal positions and the inability to "qsy hf +5" temporarily to avoid an interfering signal.
The introduction of synthesized local oscillators in transmitter drive units changed all of this overnight. The GK203 mentioned by Radio Wrangler had a setting accuracy of 3Hz at 30MHz and could be expected to stay on frequency +/_1Hz for long periods. So now it was possible to tune the receiver without having to wait for the transmitter to come up, simply by setting the channel frequency, in the confident expectation that the transmitter would be on channel to within a few hertz - good enough for everything except the most demanding of compression and multiplexing systems.
The logical outcome of all of these improvements was, of course, the conversion of most point-to-point transmitter and receiver sites to unmanned operation, all of the frequency setting, antenna selection and service selection activity being achieved by remote control from the traffic office. Such is progress.
cheers
Peter G8BBZ