It would be interesting to try a pressure pulse ... maybe generated by an air tank opening a butterfly valve.
The relevant question would be whether the degree of amplitude attenuation of the pulse would be too great for the maximum distances that you require.
The implementation of a pressure pulse detector would be relatively simple, using a pressure transducer, and some means of determining the time interval between the initial and return pulse.
It would seem that an air pulse might be relatively immune to minor hole diameter anomalies ... This suggestion would be subject to some degree of experimentation.
The argument for an air pressure pulse, used as the interrogation signal, would be that the line of travel of the primary, or strongest pulse, would seem to be primarily restricted to the physical central axis of the hole. Straight down and straight back up. ... Whereas, RF or ultrasound waves might be subject to multiple reflections withing the test hole, causing inaccurate or non-reproducible results.