I don't think many will hear(*) a difference in a high-rate mp3 (or even a lower rate mp3) and lossless. That's not really the point IMO. It's about having a 'master', so that if you need to convert later, you are not adding artifacts to artifacts, at that point, it may become more noticeable.
(*) - Even on my fairly high end stereo, the difference in mp3 and lossless is not readily apparent. But we know that something has been removed from the music. I've tested this by doing a subtraction of the compressed and originals in an audio editing program. You can now hear what is missing - little details, transients, tails of the reverb, etc. These little things are not obvious in a short A vs B, the algorithms do a very good job, but...