Kharn is right.
Gonna stick my neck out here and say I don't think either doppler or plain radar alone could work right.
Plain old radar measures distance by the time it takes for the reflected wave to return.
Velocity can then be determined by measuring how much the distance changes in a given period of time. The radar I learned about didn't do this calulation, it was done separately in a computer wired to it. In older speed cop radar, I'd guess there's an processor in the housing.
For this to work most accurately, the target should be coming straight at the beam. Otherwise, at an angle, the actual distance covered by the moving target will be a bit higher than what the radar measures.
You could figure the error using trigonometry. If you did it for a few examples, you would find that the more broadside your target travels in relation to the beam, the greater the difference between the actual change in distance (traveled by the target) and the distance change as measured by the radar.
If the target was traveling at 90 degrees to the beam, the 'plain jane' radar would have trouble detecting/computing more than minor velocity, because there was little change in distance.
Don't plan on using this to get out of your next speeding ticket, because if you think about it, the error causes [u]underestimation[/u] of the target's actual speed. And don't try making a sudden turn or change of lanes to get your measured speed to drop. Unsafe and unlikely to help you at any significant range.
In order to compensate for this error, you'd have to introduce technology that measured the sideways movement of the target as you panned the radar gun (kept it pointed at the vehicle).
Essentially this would mean measuring the angular change of the gun as you continued to paint with the beam.
There are ways this can be done in theory, but I don't know whether it's been implemented at the speed radar level, (I doubt it). Certainly the military dealt with this in a variety of ways, beginning WWII or a bit before.
Doppler radar is different, and I believe the velocity calculation is actually intrinsic to the radar circuitry. As the return time for the reflected signal changes (as the distance changes), the rate of that change correlates to a velocity.
However the angle problem you raised is no different for doppler than previous radars, AFAIK.
Even laser devices face the same issue. They measure in to/from line of sight terms, or there have been advances I don't know about (entirely possible).
It's been a long time since I studied this stuff. Maybe some P.O.'s, technicians, or engineers can add to the topic.
[red][size=4]P.R.K.