In case of HARM, large amount of the energy added by the rocket burn is wasted struggling to accelerate and climb in dense near sea level atmosphere, and the glide to the target would therefore be hampered by comparatively low maximum velocity and altitude of the missile at the moment of rocket burn-out.
Command models mach number variations at altitude (as seen in the database aircraft entries at the engine "Performance Details", with fuel consumption variations for various altitudes, speed, and mach number). The same entries are present in the DB for Weapons, and so for AGM-88 Harm.
It takes into consideration air density variation also for air friction, and for missiles, and after burnout in the glide phase? I don't know that. I would think yes, since dumb bombs, that are gliders, are modeled. If these things are modeled, the 100km radius noticed by jarraya at sea level should be considered realistic, excluding some bug.
If not... Well, hoping it will be updated soon.
But only developers can answer this question.
Missile kinematics in CMANO as of today are not always quite accurate. Can't remember how the HARM was but I'm going to play around with it some trying King of the Border this weekend. :) Generally speaking, such dual thrust rockets employ a boost phase which burns out in just a few seconds and a lower thrust sustain phase that usually burns out in few tens of seconds, tops. After that, the thing is a very poor glider, bleeding off energy like crazy. A sea level launch would not allow the missile to reach an efficient flight profile, and will probably leave it at relatively low velocity, low altitude situation after the rocket motor sustainer phase commences, having drastic impact on the achieved energy before sustainer phase burnout, and thereby on the maximum range it can reach.