I am an engineer, but I don't play one on TV.  Let's look at what a radiator
does, rather than of what it is made.

The "radiator" is a heat exchanger.  It's efficiency in exchanging heat from
hot coolant to ambient (100F in Mesa AZ, 60F in Yellowknife) depends on many
factors.  Rate of heat exchange in BTU per unit time per degree of
difference (Coolant - Ambient) for a particular motor and cooling system is
what we are trying to ascertain.  This is a measurable quantity.

In the limits a single tube with one fin does a poor job of exchanging heat.
A 1000 tube system with 10000 fins also does a poor job but for a different
reason.

A critical measurement is the difference between inlet temperature and
outlet temperature at a given ambient vs coolant temperature difference.

Let's assume the difference between ambient and hot coolant is 125F.  A
measure of efficiency would then be the difference between inlet and outlet
temperatures, as in the final hand the intent is to maximize this
difference.  If radiator A exhibits a 50F difference and radiator B exhibits
a 60F difference, it can be maintained that B is 20% more efficient in the
particular application than radiator A.  Only measurement will yield
definitive results.  A numerical design analysis will also yield the
difference in the static properties of the radiator, but it will not yield a
dynamic result.

There are many variables to consider.  Among them flow rates, residence
time, and coolant specific heat.  By simply measuring the radiator in the
desired application the actual temperature drop can be ascertained.  Of
course it is wise to have a reasonable airflow over the two radiators when
the measurement is recorded.

Aluminum, Copper, two tube, three tube, cross flow, vertical flow.  All have
effects on the static heat exchange coefficient.  Only a reasonably
controlled dynamic test reveals the data.

Attributed to R. Feynmann:

"no matter how elegant the hypothesis, or how eloquent it's presentation, if
it doesn't agree with the data, it's wrong"

Do the measurement and the result will be obvious to to the casual observer.

p