There is much misinformation out there on the stock antenna configuration.
The following information applies to the P2V+. I cannot speak for the other Phantom flavors; I don't own any. I suspect they may be the same.
The reality:
1) The antennas on the bird are simple coaxial LINEAR polarized sleeve dipoles (polarized vertically, because they are mounted that way.. They are NOT either flavor of C.P. (Circular Polarization)
This is a sleeve dipole:
2) I could not find any "official" info whatsoever on the RE700 antennas, so I dissected one.
These are twin L.H.C.P. "Perturbed Patch" antennas. I cannot tell if they are arrayed, or in diversity mode. I suspect diversity, as they have unequal feed cable lengths. The "shadow" of the radiating patch can be seen in the photo (it is between 2 layers of G-10). The upper-left and lower-right corners of the square have been trimmed (or "perturbed"), resulting in L.H.C.P. Results of trimming corners can be seen in the accompanying illustration. A patch antenna off this type typically results in 7-8 dBi of gain, depending on the ground plane.
3) A theoretically perfect Isotropic radiator would radiate equally in all directions, resulting in a spherical radiation pattern. This does not exist; however it is used as a baseline, much like 1G is for acceleration. When you "squeeze the balloon", you take energy away from some of the sphere, and project it in another direction. The resulting bulge has more gain in that bulge than an isotropic radiator, thus the gain numbers are indicated in dBi, or decibels above isotropic. The more you "squeeze the balloon", the more energy is redirected, and the higher the gain becomes. However, the "beamwidth" of the antenna becomes narrower and narrower as the gain increases. It becomes more and more "directional", Thus a high gain antenna will have very poor performance on the back side. Think of a candle flame as an isotropic radiator, and a flashlight as a high gain antenna. The candle illuminates equally in all directions very close by, while the flashlight illuminates to a great distance ahead, but virtually none to the sides and rear.
A linearly polarized antenna will radiate however it is oriented. Horizontal, or vertical, but not both. If the receive antenna is rotated 90° from the plane of the transmit antenna or "cross polarized", the gain theoretically drops to zero, although in the real world there is come cross-pol component along with reflection components. This is why linearly polarized transmit and receive antennas are never used in telemetry for aircraft. If the transmit and receive antennas are both vertical, and the aircraft rolls 90 degrees, The reception theoretically goes away altogether.
If we transmit with a circularly-polarized antenna, there are now equal vertical and horizontal components. Thus, if the aircraft with the vertical antenna rolls 90°, the gain will be the same.
I will not get into the intricacies of Left Hand versus Right Hand C.P. ,but basically they both project a "corkscrewing" signal vector out from the antenna, rotating in opposite directions.
The advantages of using C.P. on both the transmit and receive end is that any reflections are sent back corkscrewing in the opposite direction, which cannot be received by that antenna. This is why C.P antennas of opposite pols will not work together. You can't mate a left handed screw with a right handed nut.
Aircraft manufacturers, dji included, are bound by size and weight restrictions. This is not as important on the ground, which is why they sometimes go with C.P. on the ground, and L.P. on the aircraft.
You can receive a C.P. signal with an L.P. antenna, but best results will always be achieved when both Transmit and Receive antennas are C.P.