IMO BHP is a better figure to work with.
With wheel HP, you measure the power reaching the road and thats it. It doesnt take into account the losses thru the tyres or the drivetrain. As such you can see a huge increase in WHP on the same car, just by doing one test with the tranny stone cold and the tyres at 20psi, and another test with the tranny up to temperature and the tyres at 40psi.
With "Flywheel" HP, the dyno attempts to measure these losses by allowing the car to coast down at the end of the run, and measuring how much drag the driveline is placing on the rollers. This figure is then added to the measured wheel figures to give an approximation of the engines actual output.
Another example: The majority of losses thru the drivetrain actually come from the tyres (this is why 4wd cars appear to have over twice the losses as 2wd ones, theres 4 tyres being measured, not two). As such if you do one run on a wheel power dyno in 4th gear, "tune" the car, then take the second run in 3rd gear, because the road speed is lower there is less friction/drag, and you get a higher figure in 3rd gear even if you've changed nothing.
Again a coast down run would measure this difference and cancel it out.
The 4wd issue is the biggest one with wheel power dynos as at first glance two identical engines, one in a FWD car and one in a 4WD car would show the 4WD putting out a LOT less power. In reality on the road though, the two unmeasured undriven wheels of the 2wd are putting the same drag down as the extra pair of wheels on the 4wd car, so the only difference on the road is a tiny amount of losses thru the extra differentials.
A flywheel dyno would show both cars making the same power, and would more closely approximate how they actually perform on the road.