RR_NDB
Inefficient and ineffective are two different ballgames.
As I pointed out inefficient does not matter at HF frequencies most used for aircraft unless the antenna is extraordinarily inefficient. The antenna matching elements are not that bad. (If they were they'd break down from the heat they generated.)
On transmit it matters a little. But, if communications works over the required ranges with required reliability, what matters the transmit inefficiency?
Now let's look at how inefficient the antenna may be. I believe the HF antenna is in the leading edge of the "plastic" vertical stabilizer. That makes it about 35' long or a quarter wave at about 7 MHz. So it will be quite efficient at 7 MHz or a frequency near that. (It may not run the full length of the VS leading edge.) The matching circuit would have less than 3 dB of loss even at 3 MHz. Matching might get "dicey" at frequencies near 14 MHz. But even that can be dealt with. End fed half wave antennas do remarkably well with a suitable matching circuit. Zeppelins used this form of antenna, hence that antenna's name, Zepp.
3dB loss is no big deal when power levels less than a watt are known to work over distances of 1000 miles at night on 75 meters unless the propagation goes REALLY sour, which it does during periods of solar minimum. That's why God made 160 meters, one might say. If dummy load leakage can make it from Idaho to Los Angeles quite legibly under quiet conditions that gives you an easy 1000 times more power for punching through the crud Ma Nature throws at you. Planes do not use 5 watt handi-talkie power levels. They run 100 W to very conservative 1 kW depending on the radio selected.
Now let's head bang on the antenna configuration. The trailing wire can be a really nice antenna with wonderful range, if you want to talk to somebody off to one side or the other of your plane. It's not "ideal" for talking directly ahead or to the rear. Drag is horrible, especially at Mach 0.8 type speeds.
The slightly newer style of a wire running from behind the cockpit to an attachment on the tail is required for tin cans. Since the active element is very close to the counterpoise the efficiency can suffer dramatically. And it has high drag. It also suffers from some of the same directivity problems. They'd not work all that well to the direct front or rear. Finally the close proximity to the ground plane makes the antenna pretty much of an NVIS device, Near Vertical Incidence Signaling - short range regional. It hurts the long distance capabilities.
Embedding it in the tail is rather nice as a configuration. It is near vertical with, as a result, a nearly circular low elevation angle radiation pattern. In other words it is a really good long distance antenna. It works well in all directions including the critical direct forward and rear directions. It suffers a little for lack of a nice ground plane unless some screening is included around the base of the antenna. I'd imagine they've worked out something. On the whole 50% efficient is probably good enough. (Demonstrably it is good enough for transmit. And for receive it's efficient enough the inefficiency could even be increased to minimize sensitivity issues in a very high noise level environment.)
So, the antenna is effective. And it costs less than the longer antennas. So arguably it is more cost effective than the alternatives and not nearly as bad as a 9' whip on the back of an automobile. Now THAT is what I call inefficient.
Now that I've slammed this one into the ground (Earth for you British if you want the intended pun) can we quit carping about the antennas? I don't want to threaten dragging out various handbooks and playing pedagogue WAY off topic to prove my points. (The RF I am sure about. The precise VS dimensions and antenna dimensions within the VS structure are rough based on what information I could find in a 3 minute search using Google. Tail height is 9.3 meters. It's swept back some so I gave it a few more feet for the length of the leading edge.)