This is only true if the heating happens quickly and the system is less efficient when heating quickly. Otherwise, this doesn't make sense from a physics standpoint. A temporarily lower temperature differential means less kWh of heat lost.
FWIW I run my heat pump intermittently and with locally-smart TRVs that get to call for heat centrally, and a weather compensation only flow temperature curve, and it WORKSFORME!
Certainly feels like I'd need PhD from it to successfully install, modify, calibrate and run the installation until its fully adjusted, for the peak comfort and minimum cost/dirty energy use.
The obsessives amongst us can continue to tweak if we wish...
This topic comes up anytime thermostats and heating are mentioned. The physics arguments only makes sense if you don't care about comfort. Most people would rather optimize for comfort with some energy/cost savings if possible and the physics folks seem to not care about comfort at all.
The losses are proportional to the temperature differential between outside and inside.
So you should have somewhat higher losses from the hotter air streams from the radiators passing the windows.
Dunno about magnitude though.