I'm not really certain how you can explain the loss of one just one frequency it without using techie talk. At least he didn't mention that it's standing waves reflected from the break/kink/whatever that does what he describes.
What you really need to test a coax run is a time domain reflectometer. Or, if you can't afford one, look for breaks (discontinuities), kinks (impedance changes) etc, with dodgy connectors being the most likely cause.
One description of standing waves in transmission lines
HERE. Whether the reflected signal boosts or cancels the signal depends on the length (in multiples of wavelength) of the trans mission line and the signal frequency.