I highly doubt that. You're not listening to electrical voltages, which may change abruptly 48000 times a second (they don't, read on), you're listening to a loudspeaker trying to reproduce them. A loudspeaker membrane does not change position abruptly 48000 times a second. It simply can't, due to mass and mechanic and magnetic inertia. The required electrical energy to precisely position the membrane would be outrageous and instantly drive the magnet into saturation instead of achieving zero-time positioning.
Then, there are no electrical voltages changing abruptly 48000 times a second to begin with, as the discrete sample values are played back through a filter at over 20 khz which takes all the jagged edges away.
This discussion feels a bit out of place in 2024, as these myths of digital audio are being debunked for nearly 40 years now.
One thing you mention is actually true, sample manipulation (time-stretch / -compress, pitch changes) can be done with better results when there is higher resolution and thus more data points to work with.