Comparing things "regardless of details" is a pretty useless
exercise here, as in many places.
Would you call fixed frequency operation a "detail"? The
RF "client" won't.
Efficiency is the go-to differentiator. But efficiency how
and where? The marketing number, which is peak? Or a
minimum efficiency across the user's load-span (which
again is "details", so maybe you pick a converter that's
a percentage point better in the datasheet, but at your
real load-point is inferior because it's really peaky.
With a little effort you might be able to come up with
"housekeeping", conduction and switching losses from
datasheet entries and/or curves. Then Pdiss boils down to
Vin*(Ihousekeeping+fSw*Qsw)+((Vin-Vout)*duty%H+Vout*duty%L)*Iout
more or less, and
eff%=100%*Vout*Iout/(Vout*Iout+Pdiss)
neglecting things like ripple current and external losses
such as input filter ESR, output inductor R, etc.