Considering the significance of olfaction in taste perception, it seems the author is contending that the observable flavor is actually derived from the aroma(s) of the smoke emanating out of the bowl, rather than from any smoke being drawn into the mouth via the stem.