Fair enough, a global mapping of the Moon would be a useful dataset. Useful science will result from human-oriented precursor missions.

But, there will always be tradeoffs. Put in an IR mapping spectrometer with ultra good spectral resolution for global mineral mapping, or, bring along a better narrow-angle visual camera to take super detailed images of potential landing sites?

The tradeoffs were much more stark in the 1960s. Believe it or not, the Lunar Orbiters used film cameras; they could only take about 150 exposures each (five total spacecraft, one of which partially failed). The high-res coverage was overwhelmingly directed at Apollo landing sites. Coverage of the farside was really quite poor, and lots of features of potential scientific interest were never photographed. Until the Clementine mission in the early 1990s there were many areas of the moon we didn't have good images of.

Likewise, six of seven landers (Surveyors) were sent to mare sites to focus on Apollo landing preparations.