In the 1967 British TV serial 'The Prisoner' this was the question that destroyed the then super modern computer..........
I can't quite understand why you can't simply buy a normal micro cassette recorder, record Morse at 2.4 , play at 1.2 , whilst this would be far less cost effective than your idea, a '1.7volt motor' supplied at 1.2v, even going to basic e=mc2 could not run and pulse feeding would come up with the same basic result = no run.
The idea of say supplying the '1.7v motor' with 4v might work - for a while before the smoke drifts up. I think you'll find your 1.7v motor is used to (and designed for) 1.5v average supply volts so can run at 1.2v or 1.7v with only very minor speed variation of the 4.75 cms per second - in theory for half speed to the tape you'd have to be able to run it at about 0.7volts ! - Not on old chap. Though I assume you have experimented with both series feed and parallel shunt resistors the one decreasing the current the other the load - this ploy was used in the 1940's to 60's by reel tape deck designers to reduce (considerably) volts to take-up motors on play only, switching out the high wattage resistor when 'rewind' in operation.
No, I have no experience of micro CD recorders, but the use of a micro cassette recorder then perhaps feeding a standard cassette recorder from earphone jack via paralleled capacitor (.01mf) and resistor (about 1kohm) per channel into line input of standard cassette recorder , making master tape then duplicating on a high speed dubbing double deck unit (which has the effect of slightly increasing the upper register of the sound) might be the answer to supplying a whole class with off-air recordings. I do so agree the lab-tutorial Morse tapes are hellish boring.
I do not agree with the observations on vinyl disc or even shellac 78 rpm record play quality (crackles not withstanding) - it does very much depend on the quality of pick-up, player, filters/amplifier and size/quality of woofers and tweeters speakers. Please remember this was analogue state of the art then - one day folks will look back on our discs (CDs DVDs etc) and smile as the EEPROM still has a lot of development before truly solid state record/playback of any time duration and frequency response can be achieved digitally. Best wishes.