I have been reading David's keen insight for quite some time and have learned a lot about many different subjects. There is no "making stuff up as they go along" with him. And I do agree with the title "resident expert" because that is exactly what he is. Thanks for all of the education you have provided me free of charge, David - it is appreciated by this industry person.
If a song was a "deep cut" in 1970, it was not popular enough to be a hit and therefore will not be included in a "classic hits" format. In 2014, it is even a deeper cut, and when all music research is considered, only the true hits of the era will make the cut. I look at it as you are only as good as your weakest song, and deep cuts that didn't chart are going to be pretty weak to the average listener.
I have had this argument with a oldies DJ at a station I used to manage. He wanted a 4000-5000 song library and we were only using about 1000 songs, which is likely a stretch in reality - 800 would have likely been better. He didn't understand that while he may have known 5000 songs, he was the exception and was in the "1%". He listened around the clock and claimed to hear repeats too often, and refused to accept that the average listener only listened for about 45 minutes. His attitude was a good reason to never let him be the PD as he had zero understanding of programming to the masses. Familiarity is what the masses generally want, and music test after music test will confirm that.
If you had a 1000 song classic hits library and you played 14 songs per hour, it would take about three days to play the entire list. If the average listener only listens one hour per day (which is a stretch), they may not hear a repeat for months - yes, months. And when they did hear a repeat, it wouldn't feel like one as it had been a significant period of time since they heard it last. I know that may be over simplifying it, but the numbers rarely lie.