There are to questions in CFAI curriculum where they use two drifferent denominators in divident discount model in terms of REITs terminal value calculation (reading 39), namely:
the schweser errata corrects both instances that they use re-rfr for the terminal value to re-g.
I’m slightly concerned that it isn’t in the cfai errata as i’m sure someone would have informed them. I still can’t think why you would use the required return minus the risk free rate though, it makes no sense.