In CT, detection range is something like 300,000 km while tracking range once detected is like 900,000 km. So Fighters positioned 300,000 km from the planet can locate the Deep Meson Gun, while taking fire from it ... yet the FLEET can fire on the planet, DMG and enemy Defense Force from ranges that deny the planet any chance of detecting the fleet to get a target lock on it. ...
Detection is two light-seconds, but that's a quibble. More on point, how precisely are the fighters detecting the meson batteries?
On a semi-related note, I wouldn't as a gamemaster allow missile fire on a planet-side target in atmosphere from three light-seconds off. If missiles could survive re-entry at those speeds, they wouldn't even notice laser fire. Conversely, if they slow down to manage a re-entry, they're sitting ducks for defensive fire from the planet. Even if they could, having a ship pick out a specific target against the surface of a planet from three light-seconds off is not the same as picking out a target against the background of space. Always figured the reason
Invasion Earth required ships on bombing runs to be in close orbit was because the missiles either wouldn't survive re-entry, would get shot down if moving slower from farther out, or the ships wouldn't be able to distinguish targets against the background of a planet's surface from farther out. I also wouldn't let a ship get away with firing weapons while remaining undetected, even at three light-seconds. But, other people may have different views.
And, from Striker Book 2: "The most common form of active defense in the deep meson gun site. A deep meson gun is a meson gun of ship ordnance size buried in a deep underground chamber. As the planet itself is transparent to the meson beam, the meson gun can fire at any target desired, while the site itself is effectively impossible to locate. Only when the gun site's surface sensors and target acquisition devices have been destroyed or captured can the gun be silenced, this generally requiring the use of ground troops or extensive planetary bombardment."
MegaTraveller followed through on that theme. EM sensors aren't going to detect something buried deep in a planet, and densitometers go to 1 Km at TL15; below that depth, you couldn't detect anything except neutrinos, which means the power plants are vulnerable but the buried guns aren't.
Several points here:
The MG sites are not,as you say, invulerable, at least to meson fire from space, as the MG can pass through any matter. The main point here is detecting them with enough exatitude to fire them (though probably a near miss would have some effect, be it due to EMP, radiation, damages on power gird or collapsing of part of the tunnels. ...
See response above. One cannot hit a target if one has no idea where it is. I can argue for eroding the power supply, and I'd have to develop rules for locating and destroying sensors if I were inclined to model such a combat, but the gun itself is pretty much impossible to find unless your intelligence agents managed to learn its position.
... OTOH, as most planets in OTU are relatively low population ones (among hundreds of thousands and a few millions) and lower TL than the Imperial troops, the planetary invasions are a true possibility with relatively few troops, and their capacity to build DMGs is limited.
Depends on tech level, which model you're using for GWP, and a couple other starting assumptions, but I'm getting in the vicinity of 300 thousand or so population as the minimum to afford a single installation at TL13. At that it's eating up 15% of GWP annually, so it's most of a planet's military budget. Less at TL15 because of the higher per capita GWP and cheaper power. Gets rather expensive before that, between the lower tech systems and the lower per capita. Mind, it's a good investment if the point is to keep people from threatening your cities with nukes from orbit, but it's a pretty heavy tax burden; whether it's worthwhile or not is likely to depend very much on local circumstances.