I think these theories are general. So you can use them in any freq. range. The feedback is more offen in schols, but in my opinion negative resistance is more practical.
It is correct that there is only a very thin line between the two. Often so called "negative resistance" topologies can be rewritten in a feedback format by introducing a virtual ground node different from the real ground node.
Fact is: an oscillator should only do one thing: oscillate. The way you analyze the circuit is of less importance.