Karl comments on Rational vs. real utilities in the cousin_it-Nesov model of UDT