Apple Says It Doesn't Know Why iTunes Users Are Losing Their Music Files (theverge.com)
Earlier this month, an Apple Music user James Pinkstone claimed that the online music streaming service deleted 122GB of music from his library for no apparent reason. Several Slashdot readers noted they had also faced a similar issue or knew someone who had. The iPhone maker has now acknowledged a bug in iTunes that is apparently causing the glitch, however, it adds that it doesn't really know why "some" users are facing this issue. The Verge reports: The company confirmed that "in an extremely small number of cases, users have reported that music files saved on their computer were removed without their permission." However, Apple was unable to reproduce the bug, indicating it doesn't really know what's going on here. The company adds: We're taking these reports seriously as we know how important music is to our customers and our teams are focused on identifying the cause. We have not been able to reproduce this issue, however, we're releasing an update to iTunes early next week which includes additional safeguards. If a user experiences this issue they should contact AppleCare.
You've got it mostly right, but not completely right. Any sort of digital manipulation introduces noise into the audio stream due to the nature of floating-point math. The more accuracy (or "headroom") you have in both domains (frequency and amplitude), the closer the final mix will be to the original waveform. That's why many studio engineers use 24bit / 192khz source material... not because you can actually hear any difference in the original source, but because it preserves the pristine quality of the source longer as it moves through arbitrary numbers of digital filters and mixing stages.
It's moderately useful for masters as you stated, because those masters can still be digitally manipulated without introducing any additional audible artifact. As indicated by vel-ex-tech, you actually do need additional samples for clean pitch or tempo shifting, but I wouldn't call that a typical "consumer" application.
Of course, for consumer tracks (simple listening), it's utterly pointless, much like the ridiculous Pono player. No one can actually hear the difference, and nearly all reputable double-blind studies confirm this. It's just a "bigger numbers are better" fetish among audiophiles or people who don't understand the science of audio. And frankly, the "loudness wars" have actually decreased the require dynamic range for most consumer music by pumping up the volume and compressing the shit out of it.
Irony: Agile development has too much intertia to be abandoned now.