Longer antennas (properly designed of course) focus your signal so it's concentrated towards the earth's surface. We're not talking to airplanes or other planets so why waste signal up there. We want to focus our signal at ground level (or horizon) which is where our contacts are.
The gain of the antenna tells you how much it concentrates your signal where we want it.... in this case the horizon.
So gain (length) and height both help.
You may also need to consider the loss in your coax. You may want to run heavier and lower loss coax for the longer run then adapt to something more flexible for the last 10' or so. I would try and keep coax loss down around 2dB or less if you can. Less than 1dB would be nice.
In a lab environment, when they measure an antenna to derive its radiation pattern, they reference the center of the radiating element(s). So it would seem logical that if you were doing scientific experiments with antenna and wanted to compare the coverage of two antennas in real-world space, you too would reference (or mount) both antennas at the same elevation, using the center of the radiating element as the reference elevation. For simplicity, let’s say the smaller antenna had 2 foot radiating element and the larger had 10’. If you mounted the 2 foot antenna at 45’ AGL to center of radiator and wanted to contrast it against the other, you would mount the 10’ at 45’ AGL to center it its radiator.
For the sake of a fictitious example. Assuming that the gain of the 2’ and 10’ were both the same (they would not be), the power used was the same, and both were mounted at 45’ to the center of the radiator, I would expect the shorter antenna to win the range contest because its center of radiation is higher above the horizon. Whereas if you compared them both by mounting the bottom of the radiator at 45’, now suddenly the 10’ would win because its center of radiation is higher than the smaller.
Now let’s switch to reality. For GMRS, the longer antennas have more gain because they are in-fact some form of collinear antenna design. They are actually comprised of an array of multiple antenna stacked one on top of the other. This has the effect of increasing the effective antenna gain (squashing the donut as some would say).
So, if you have an antenna mast and it is 45’ off the ground and achieve range ‘X’ when using a simple ground plane antenna, you will achieve range ‘Y’ when using the 10’ collinear antenna. ‘Y’ will be greater for two reasons. First, it is a higher gain antenna. Second, because the center of radiation is higher.
If you live in a valley and need to communicate with radios on the ground in the valley as well as others high on the nearby hills/mountains then a lower gain antenna is usually better. If the terrain is mostly flat to rolling, than a higher gain antenna becomes a candidate.
One finally thought. If you live in an area with lots of trees (as I do), a low gain antenna mounted well above the trees is probable to substantially out perform a high gain antenna located below the tree line. So while we say that height is everything, which is true, what is is really all about is getting the antenna above all the obstacles (trees, buildings, earth, etc...) that reduce its range.
