As for me, I cook a lot. The palettes of scent and taste are important and the vocabulary of spices and herbs and major flavours is essential.
The same logic goes for smoking tobacco. I can very much see the sense in trying out base or blending tobaccos. It helps inform your palette just like knowing ingredients in a good dish. I can sense many flavours in any dish due to my knowledge of spices, herbs and signature flavours and scents. So like others have mentioned, smoking base tobacco types would make sense to build that knowledge. If you do so, smoke it in a pipe that won't colour the taste (like clay or meerschaum - I find new cobs tend to add a bit of corniness to the profile)
With tobacco, there are so many different ways of treating the same leaf. It really helped to try blends that were very heavy toward red virginias or golden. Blends that contain Burley or fired Kentucky. Blends with more 'turkish' or oriental leaf, or Latakia. For blends that had a lot of cavendish ~ I've found this really varies a lot by producer, I'm guessing that's the variation on how they produce their cavendish and what topping or casing they apply to it.
Compressed tobacco's with the same leaf combination as loose leaf blends can be wildly different as the oils ferment and mix completely differently.
The technical side of how you smoke anything also changes its flavour (smoke slower is always good advice). So taking notes is very useful. If I smoke a straight Virginia flake of one brand (for example SG Virginia Flake) it does taste different if I rub it out and roll it into a very loose ball as opposed to folding/stuffing or cube-cutting. I find that I develop a favourite method with each blend after trying things out. Sometimes I get an overly mild or boring smoke from folding and stuffing, but rubbed out the flavour shines more.
Smell and taste are inextricably connected. I also tend to avoid reading reviews of something before I try it - I don't want it to influence my own views. I'll make notes and read them afterwards though.