Bit perfect audio is certainly a desirable goal, but it is not the only consideration in playing back digital music. The whole D to A conversion process is not a closed end solution. That is, it is an approximation to the exactness of the Nyquist (or Shannon) Theorem. Therefore, DAC manufacturers have adopted multiple methods of D to A conversion, almost all of which include a lossy transformation of the data into a high sample rate, low bit form. This is the well-known Delta-Sigma method, which is designed to simpify the playback circuitry. In fact, the actually signal sent to the end stage circuitry of most PCM DACs is more DSD like than PCM like.
Some people think that a properly designed DSD DAC sounds different than a PCM DAC, just because of the nature of the processing. Then the questions is whether the lossy conversion of PCM to DSD is worse than the advantage of DSD playback. For me, the answer to that depends on the details of the DAC processing and the preference of the user to different types of sound. And, yes, different DACs do sound different to some people.
I totally agree that lossy conversion should be generally avoided. But, it should not be a hard and fast rule, given that some people honestly like the sound of DSD better. Likewise, a statement that DSD always sounds better than PCM is not appropriate. The individual's preference of the sound really needs to be considered.
I assume some will disagree with this assessment, as we have have been down this route before. So, let's not get into yet another long discussion that leads us nowhere. I just wanted to get an alternative view on the record.