I don't lie to undergrads, however it is my responsibility to educate them properly by showing them where the data comes from. These values do help explain phenomena in organic chemistry.
What's more reactive? An acid chloride (conj acid pKa -8) or an amide (conj acid pKa 36-40). The pKa information is necessary, and standardization is necessary, and providing student with some inkling of where the data comes from is necessary. There are other arguments, however I prefer not to switch arguments when it's convenient.
It appears that crossing solvents requires some interesting mathematics:
http://lisa.chem.ut.ee/~ivo/HA_UT/And the mathematics create errors as high as 1 log unit (minimum). In the past 30 years I've seen the values of pKa data for NH3 vary as much as 5 log units! Not trivial.
The more I think about this problem, the more I want to turn it into an O-Chem lab assignment... once I understand it myself.
Thank you for your kind comments.