No, it was PD. We're interested in characterising precisely what PD's "flat-floors" are supposed to be, since they don't behave like flat-floors (alone). You did agree with the golf ball analogy, though.
Why the attitude? What I said is true. Read
here.
Making an object "rough" does reduce the aerodynamic efficiency of the surface flow layer. In some cases that can be beneficial
overall by bumping the object into the turbulent regime "earlier". But, crucially, if a golf ball were bigger and / or traveled faster, or if it were smaller and / or traveled slower, the dimples would just increase drag as normal or have no effect, respectively. It's just a happy coincidence that golf balls sit in that sweet spot of Reynolds values where you can exploit the effect, and was discovered by accident (and later verified experimentally and theoretically).
The reason the golf ball was mentioned in the first place was to say that a rough surface is more aerodynamic, so a flat floor should create more drag. However, a flat, smooth floor on a car will always have less drag than an equivalent "rough" one, all else being equal, because a car is generally in the fully turbulent range when aero drag becomes an issue (Re of 10^7). The "regime-bump" won't work in that case, so the comparison is invalid.
I think you're just picking at details for the sake of it, and it's getting in the way somewhat.