For a start, sampling at 98kHz is pointless. The highest frequency you could theoretically hear is when you were a kid was ~22kHz, but realistically, if you're past your teens it's below 20kHz and reduces with age from then on. So the minimum sampling frequeny for no loss of information would be 40kHz. 44.1kHz is the CD standard - a frequency chosen because of the early use of video tape to record audio and it was the lowest common frequency of the useable 525 and 625 line frame and line rates of broadcast TV. It also gives a bit of headroomn over the 40kHz theoretical minimum for anti-aliasing.
The encoding bit depth has nothing to do with the sampling rate - it just sets the dynamic range of the recording. Going from 16 bit to 24 bit increases the dynamic range from 96dB to 144dB. Some audiophiles with perfect setups, no neighbours, a soundproof room and trained hearing might appreciate the extra range but the other 99.999% of listeners are using iDevices with bluetooth headphones or crappy earbuds and it would be wasted on them, so why waste the memory and processing?
Finally, where are you going to get these recordings? CDs are recorded in 16 bit, so ripping them at 24 bit gains you nothing whatsoever. The theoretical dynamic range limit for perfect vinyl is about 111dB - but only the first time you play it. The old classical master cuts probably had close to 100dB when they were new - but most vinyl you bought in Virgin records would have been down at 75dB even when new.
TLDR - because from a technical perspective the differences that higher sampling rates and larger dynamic ranges would produce would be inaudible to the vast majority of the people who listen to music on their iDevices with a resultant increase in file size and processing power requirement which would offer no tangible benefits.
