Couldn't agree more, Morbid. I have several times extended an offer via this BBS to HtC to act as an unpaid consultant to help them build a higher fidelity radar model into this sim (Graduate degree in Radar Signature Prediction and Reduction from the US Air Force Institute of Technology). I made the same offer to iEN -- they weren't interested either;(.
What I would suggest is a fairly simple model, where detection range is dependent on aircraft altitude (i.e. line of sight, as modified by ground-clutter), aircraft size, aircraft aspect angle, and terrain masking. A couple of standard radar signatures would be used, one for single-engine fighters, another for twin-engine fighters, another for medium bombers, etc. These signatures would consist of five values, and the model would determine which one to use depending on which cardinal direction was facing towards the radar antenna. In other words: nose-on, left/right front quarter, left/right broadside, left/right rear quarter, and tail-on.
The host or FE would determine the type of target, the distance and altitude of the target, and determine the aspect angle of the radar relative to the target. It would look up the RCS (radar cross section) for the target type and aspect angle, and plug this value into the basic Radar-Range Equation to determine if detection is even possible. If it is, the software model would then determine if line of site exists, and whether or not the target is in ground-clutter or not. Each FE could perform these functions, and simply send a "yes" or "no" to the host; "yes, the radar sees me" or "no, the radar doesn't." Obviously, a radar that has been destroyed would not be able to see an A/C; but then, your FE would know if a radar was destroyed.
This is a gross simplification of the actual electromagnetics involved, but would be far better then the current model...or lack of one. The biggest potential problem with implementing this is the potential to hack your FE to change the RCS tables. This could be avoided by making the host do all the work, but might bog it down too severly. You could also play around with the update rate, making it longer or shorter, or even range dependent; i.e. the farther you are away from the radar, the less frequent the updates. I also agree with the suggestion of reduced icon viewing range of low-flying a/c. What do you think?
Sabre (a.k.a. Rojo)