Yes, vinyl can indeed sound different, but this does not automatically indicate a limitation of the digital format. The digital didn't lose any information itself if a mastering engineer took that information out - the digital was just given less information to work with.
The brief summary (for anyone who cares):
For non-lossy-compressed digital, the most obvious loss of information is high frequencies that need to be filtered out and anything else that gets thrown out in the process. It is indeed a limitation of digital formats that frequencies >= half the sampling rate need to be filtered out.
The other way you can lose information is if you don't dither and reduce bit depth then you truncate some low level information that might otherwise be audible in the noise.
After that we are talking about stuff below the noise floor or otherwise masked by noise, distortion, aliasing or any other artifacts.
Given the noise floors of vinyl vs. CD quality digital, the most likely reasons something would be audible on vinyl but not on the digital side is that something happened to make the audio different beforehand. It could have been a hotter recording (as you suggested), EQ, compression, a different mix or any number of other things done for any number of reasons. I suppose it might be possible that various vinyl distortions could make something more audible, but that's hardly a limitation of digital.
For lossy compression, psychoacoustics are used to attempt to throw away stuff that the ear is going to throw away itself before it even gets to the auditory nerve, much less the brain. To the extent that the encoder is successful at this, even though it's "lossy" it can indeed be transparent to humans. Modern encoders at higher bit rates are quite successful at this, as anyone who has done a proper ABX test knows. There are some signals that are more difficult to compress transparently and there are some people who are quite good at hearing artifacts, but the problems with lossy compression are often vastly overstated, especially at higher bit rates and especially with people who don't understand that their own ears are "lossy". The easiest way to evaluate for one's self how well this does (or doesn't) work is to just do some ABX tests and see.