They can disagree all they want, but they’d be wrong (data from Ahlquist et al.).
There are also studies showing that cadence has no impact on the mean frequency of the mechanomyographic signal (also indicating an unaltered pattern of motor unit recruitment).
Ditto with respect to EMG (except in the rectus femoris) (cf. Figure 2).
And lest anyone forget, not only does pedaling at a slow cadence not result in selective type II motor unit recruitment (vida supra), training that way doesn’t seem to improve performance (vida infra).