high quality speakers are required but mastering from higher bitrate and word length is generally going to yield noticeable improvement.
It's a no brainer when the price of the song is the same.
I'm afraid it's not a no-brainer because contrary to what you seem to think, people actually PREFER compressed dynamic range in most tests (i.e. the whole mentality of 'louder is better' right down to LP Vs. CD even). Higher samples rates are utterly useless (we can't hear over 20kHz so it's just wasted bits) and your average recording doesn't even have 12-bits of dynamic range, let alone 24-bits so again, it's meaningless in most cases (20-bit as about as far as you can POSSIBLY hear and that's going from barely audible to Space Shuttle launch in an instant, which thankfully most recordings do NOT do or we'd all be deaf if we had it turned up where we could hear the quiet bits).
24-bits is great for recording because of the excess head-room it affords. It's NOT NEEDED on the consumer end of things. 18-bit would be more than sufficient for nearly any recording ever made and good old 16-bit audio is more than sufficient for 98% of all recordings out there. There is some possible benefit to multi-channel sound, however, but I've never seen any on iTunes either way.
And before someone rants on about how they can prove how much better various SACD or DVD-Audio recordings sound, just keep in mind that many of those recordings are remastered first. In other words, the reason they sound different and even much better in some cases is that they used a good quality mix for once instead of that compressed CRAP they put out for radio (studios are notorious for mixing for the least common denominator and many use crappy speakers in their studios for just that reason). Dump a 2-channel mix of SACD to CD and you'll find it still sounds virtually identical.
All of this is easily proven with DBX switch boxes. For those that doubt, look into the actual scientific tests on the matter.
Meanwhile, my own recordings in Logic Pro sound exactly the same when dumped as WAV as 256kbps AAC here. It's audibly transparent to my ears and I have quite a high-end ribbon speaker setup with a custom active crossover and bi-amping.