Well, it's a little deeper than that.
Counterpoise is the incorrect term, but lots of folks use it. Ground plane really isn't the right term either, but it is the most common. The real issue is image plane, but we're more or less dealing with the same animal.
Whether a vehicle is an adequate ground plane depends on the frequency. For a 70 cm antenna, it's probably adequate, and to a lessor degree it's okay for 2 meters. However, as you go below 2 meters, the ground losses mount up. Depending on the actual installation, the ground losses on 6 meters are about 2 to 4 ohms, total. This includes the actual ground loss as well as a little for the mounting method.
On the HF bands, the recognized ground loss (calculated) is about 2 ohms on 10, and about 10 ohms on 80. However, the real-world loss is about double these figures.
The original (accepted) calculations were done by Dr. Jack Belrose, way back in 1953. About a year ago, the good doctor postulated a new rule as it were. Where in, you measured the antenna's input impedance, and then modeled the antenna. You then adjusted the ground loss figure in the modeling program, until the two were equal. Thus, you'd arrive at an efficiency rating for your installation. If you go through the exercise, and you model your installation correctly, you'll discover the losses are as I suggest, about double the calculated losses.
Remember, anything you put between the antenna, and whatever ground plane there is under it, adds to the ground plane losses. The only way you can compare them, is to measure the actual input impedance of the antenna in question, but that won't tell you directly. You have to make an improvement, and then measure it again, and note the change.
You just might be amazed at what you find.