satellite.targeting_coverage_factor

Crash targeting: candidate samples tested per clearance-footprint, i.e. per (targetingRadius/clearanceRadius)^2. The sample count scales with the targeting area, so a small circle runs far fewer checks than a large one. ~1.2 is bare geometric coverage (samples spaced ~one clearance apart); the default adds margin to catch tight gaps. Pushing it much higher mostly adds correlated, redundant checks.