Lots of speculation goes into antennas for portable use. There are conflicts. On the one hand you need to cart all the stuff with you so there is a definite "less is more" attitude here. On the other hand, putting up a "dummy load" does no one any good. As it turns out, height of the pole or poles available in wavelength determines just about everything in the selection process.
Several basic types are compared here using EZnec software. One is a very highly optimized vertical. It uses 16 elevated radials 28 inches off the ground. It uses lots of segments and is carefully matched to the EZnec model to make as certain as possible that it is free of computational artifacts.
The second is a classic dipole erected using about the maximum length of pole that can be easily packed or taken on an airplane, a pair of multi-segment 16 foot poles. The third is a classic inverted V with an apex angle of 90 degrees. Another was modeled with an apex angle of 120 degrees but it had a large footprint due to the long guy wires. It was a tiny bit better at the flatter angle. The final one was an interesting fan dipole with triangular wire elements that were 24 inches wide at the far end. All used the same pole height.
The shock was the vertical. First look at the great pattern!:
This looks REALLY good. It should. Everything that can be done has been done. This design brings along 16 additional stakes each about 3 feet long, to hold up the 16 elevated radials, but it is self supporting and needs no guy wires other than those on the 16 stakes for the elevated radials. Look at the nice concentration of power in the low angles, something a vertical is famous for. This one is getting close to the pattern over perfect ground.
Let's look at the pattern of the dipole. Who in their right mind would cart along two 16 foot poles instead of the above antenna. You could even get away with not elevating the above radials and laying them out on the ground as long as you used 16 or more of them, and still get much the same performance as above. But here is the dipole pattern:
Not much to look at compared to the beautiful vertical pattern above. But there is something to look very very carefully at here. The absolute values. All of these antennas were modeled on 5 millisiemen ground and in each case the support poles were modeled as well as grounded metal poles. Note that the presence of the poles made only a slight difference in the patterns and gain figures. All antennas were carefully resonated at the test frequency.
So what is wrong? Let's have a look at the two patterns scaled and plotted on top of each other:
Oops. Suddenly the dipole does not look so bad. In fact, there is only a very tiny portion of the patterns, below 9 degrees, where the vertical is better than the dipole and there only by the smallest of fractions of a db. The vertical pattern is a subset of the dipole pattern. At angles like 15 degrees, a solid DX angle, the dipole is considerably better. And this dipole is only mounted about 3/8th wave above the ground.
The fan dipole was fun. It was of course shorter between poles than the regular dipole and seemed to have a slightly better gain figure vs. the dipole. Perhaps it had a bit bigger capture area:
But notice the great similarity to the regular dipole. This is a cute antenna. It has one advantage. Since it assumes a slight droop in the upper wire, it supports the coax better than the regular dipole. The regular dipole was modeled with a droop due to coax or feed line weight. With an upward angle of 5 degrees on each side if center. This made a slight improvement in the pattern of the regular dipole.
So we decided to simplify things. With an inverted V you only need one support. Half the things to drag along. But there is a serious problem. As soon as the dipole started to droop a significant amount downwards at the ends, cancellation attacked. The sad truth is the inverted V is probably a bit worse than the vertical:
Let's compare the two on the same scale on the same plot:
Note that below 20 degrees the vertical is better, enough better to call it significant perhaps. So, the sorry truth is that in spite of how pretty the vertical pattern is, and how nice it looks, and how primitive the old center fed dipole is with its simplicity, the dipole appears to be the best selection.
So why the great reputation of the vertical as a DX antenna? Probably for what it throws away. The pattern of the vertical will make it reject signals coming via NVI and high angle skip. This means it will reject signals from 50 miles to 400 or so miles out. If you live in California this means that lots of DX signals buried in local QRM might suddenly become readable as the vertical rejects the local stuff in favor of the low angle signals.
The vertical with ground laid radials is easy to set up and self supporting. When done this way it is a compact antenna to take along. The problem comes when you try to do better. You can use a pair of verticals and phase them. But then you have the same double hardware of the two poles supporting the dipole. The final sorry result is that phasing verticals is a bit tricky for best gain, you need lots of radials under both verticals to get them to work well and you might get about 3 dbi gain. The dipole is very easy to set up and tune and get the predicted performance out of. And it may well outperform your fancy two phased verticals unless you get them just right.
The support poles can be quite light, of nesting aluminum, relying on small guy wires. The dipole itself can be designed to be a light as possible as well to reduce strain on the poles. 16 feet should be quite manageable with careful design, making a bundle short enough to qualify as "fishing rods" a category recognized by airlines.
Finally, one last look, this time at the Fan Dipole vs. the optimized 1/4 wave vertical with 16 elevated radials. The cursor is placed at the intersection of the two figures and the lower left hand corner has the gain at that point:
At one half wave above ground, a simple resonant dipole will significantly outperform the best installed 1/4 or 5/8 wave vertical.
At one quarter wave above ground, A simple resonant dipole will outperform a quarter wave vertical for NVI or local communication. The vertical if well installed over a decent ground structure, will outperform the dipole for radiation angles less than 30 degrees.
Below one quarter wave above ground, the simple resonanat dipole will suffer significant loses, but still outperform a quarter wave vertical for NVI operation. Even a shortened loaded vertical, well installed, will outperform this low mounted dipole for radiation angles below 30 degrees.
An inverted V has both problems, the self cancelation of a vertical and the low mount pattern distortion of a dipole. But at 1/4 wave center mounting height it would be a good compromise between a quarter wave vertical and a horizontal resonant dipole. It would be much easier to make efficient if one is not judicious about the ground screen creation for the vertical.
Thus selection depends entirely on the height available and the number of supports. With two supports 1/2 wave tall, a resonant dipole, especially a fan dipole, will outperform the vertical in all cases.
With less than 1/2 wave of support/vertical pole, a carefully installed vertical will outperform the dipole for DX, but be much less useful for local NVI communications.
The inverted V includes the worst of everything, and is thus generally useful with a single 1/4 to 1/2 wave center support.