This question of efficiency has been bugging me all day. I think it bears some discussion, although I suspect I will be the only one interested.
I think one of the better (sorta) lay descriptions for efficiency comes from Tom, W8JI
http://www.w8ji.com/radiation_resistance.htm
He says, in part,
Quote:
Folded monopoles provide the clearest common example of radiation resistance misuse. Quite often, in discussions of vertical antenna ground system loss, claims are made that multiple drop wires increase radiation resistance and lower earth or ground system losses. The justification is multiple drop wires, or a folded monopole element, increases radiation resistance. The increased radiation resistance reduces ground currents and ground losses. This concept is justified and/or rationalized through use of the common formula eff % = 100 * Rrad/(Rrad + Rloss) .
Losses must be normalized to the same point where radiation resistance is taken, otherwise the efficiency formula above does not work! Many folded monopole articles either ignore the fact that loss resistances must be normalized to the feedpoint, or the authors are unaware of that rule.

So total efficiency is given by several factors for physically and electrically short antennas, which if you notice means that most of our space limited antennas will have very low radiation resistance but high radiation capacitance, necessitating lots of correcting inductance to minimize antenna reactance.
The definition (nonnormalized) for radiation efficiency is given by:
 Radiation resistance Rτ
 Groundterminal resistance Rg
 Resistance of tuning inductance Rc
 Resistance equivalent of insulation loss Ri
 Resistance equivalent of conductor loss Rw
Notice that ground resistance, tuning inductance and losses all reduce efficiency and so even a physically good antenna is ultimately subject to the same factors as a poor one, which is why when you read eHam you get so many 1of5 AND 5of5 for the same antenna. Also notice that radiation resistance is both in the numerator and denominator. If you take the time to do good return work, a Hamstick has sufficiently high radiation resistance that it can match fairly closely the performance to higher dollar units. IOW, it's often the care and feeding of the antenna and feed that makes or breaks it and even a random length of wire can work pretty well with the right conditions.
Ultimately, the measure of efficiency, though, is how much power is contained in the generated field vs. input to the antenna. This is where most theoretical efficiency descriptions get sidetracked, because they assume feedpoint characteristics that aren't necessarily true. Anyway, the sure way to include all this theory is to measure power into the antenna and field strength some distance afield. When you do this you defacto normalize and can compare. To which the real world experience of rhyary starts to make sense, what he's doing is simply working pretty well!