I have the Sennheiser HD215 and I used to listen to the FLAC
music only and today someone told me not to waste money on the
CDs since 512 bitrate music is available on the net and it is not
really possible for the human ear to notice the difference.

What's your take on this?

512 means really good quality, usually the player/amplifier or the headphone itself is not good enough, noisy or .... But it also depends on your own ear.
Therefore in general yes, you will hear no difference, but there can be cases.....

I can tell the difference.

Storage space is plentiful these days, I can't see why anyone would want to run MP3...

I also can tell the difference. I also believe that mp3 was obsolete a decade ago and am unsure why people are still using it.

What's "512 bit rate MP3"?

The MP3 format maxes out at 320k/s.

Ogg Vorbis at -q10? No, I can't tell the difference. I can't even tell the difference with LAME at 320k/s.

That's what I thought as well. In fact even if you were able to encode past 320, there would be no significant benefit in quality.

Some people claim there's a difference in sound with wooden knobs on their stereo and little bags of rocks taped to the cables. It's a highly subjective area.

I'd say there is no difference in sound quality except when there is...
Most of the time there is no difference between high bitrate lossless and FLAC but, for example, I used to have high bitrate MP3 and FLAC of some of the same tracks on a device due to having changed media players and other things and I could tell immediately when the MP3 was playing of a certain NIN track (I forget which) because the modulated noise at a certain point lacked some "texture". Don't think I'm being all "golden ears" here -- I would defy anybody not to notice the difference.
However, with some tracks I didn't immediately know which version was playing and with some I might not be able to work it out at all.
So, to my mind, the safest way to go is just use FLAC for everything than I know I'm not missing out.

It also depends on what headphones or speakers you use. If you're using ear buds, there's no way you can notice any difference. I guess that may be why mp3 is still popular.

I think I was using my Shure SE535s at the time but my Ultimate Ears TripleFi 10s show it up just as well even though they're less detailed to my ear.

Also because everything supports it. My car supports very few types of music, only WMA, M4A, MP3 and AAC. So mp3 it is. Until they upgrade the head unit to support everything, I don't have a choice since it's leased, and a new head unit is not permitted during the lease period.

My car only supports CDDA, I don't mind.

As for me, I have an enormous collection of plastic discs (CDs ...), and my ear is perfectly satisfied with the audio quality. I sincerely doubt that the addition of "one measly bit" would change that opinion.

Remember that all of the digital (or, digitized ...) recordings that you're likely to encounter have been mastered to the digital bit-width of the CD. Therefore, the high-order bit(s) of any "wider" recordings are going to be ... zero. There is no "additional data (precision)" to be captured! The sounds that were present in the mastering-console were digitally mapped to ... the numeric precision of a CD.

(And subsequent "answer prints" may well have been made for MP3, AIFF, and so on. The Beatles always had "a cheap set of speakers" in the control-room at Abbey Road for precisely this reason. The speakers are still there. . .)

On playback, the numeric range of the recording ... of the data stream ... is mapped to an inclusive range of amplitudes, which is governed by whatever the end-user's sound system might reasonably be expected to reproduce. If you add, say, "one additional bit" to the data stream, the only thing that you've actually done is to double the precision of that data-stream to choose amplitude levels within that same gamut, which is harshly determined by the capabilities of the end-user's sound system (or car stereo). You haven't changed the capabilities of that sound-system at all: you've only increased your theoretical capacity to select among them ... i-f that actually matters. And, as the Beatles' "cheap speakers" confirm, it really doesn't.

"Face it, you're mastering for earbuds!"

Adding one bit doubles the amount of availeble sample values.
Today, studios use 24 bit/96 kHz digital mastering. That results in a dynamic/frequency range significantly above that of a CD. They actually have to use low-pass filters when downsampling to 16 bit/44.1 kHz to avoid artifacts, as high frequency signals turn into lower frequency noise when (down)sampled with a sampling frequency below 2x the frequency of the signal in question.

Back in the 1980s and early 1990s, 16 bit digital mastering was pretty common, but the sampling frequency was usually 48 kHz rather than the native 44.1 kHz frequency used by CDs. 48 kHz is the native sampling frequency of all early Digital Audio Tape (DAT) equipment.

I bought one of the so called K2HD cd's from Japan a while ago. Sure it is kind of a trick since the cd format can't improve upon the limits but the sales pitch was that it was re-mastered in higher quality and downsampled to fit on a cd.

Kind of sucks in my opinion. The vinyl version still sounds better.

Can ones ear hear the most minute of sound? Seems that the trained human ear can in fact be quite precise.

Can an average user hear it? I doubt it.