@Joseph Hernandez yeah this is actually a really important distinction that trips people up — the 50km figure is drone-to-drone specifically because both ends are airborne and you've essentially eliminated ground reflections and multipath, which means you're getting close to true free-space path loss conditions. At 10-15km with 60-70% first Fresnel zone clearance you're going to see a measurable RSSI hit, probably in the 3-6 dB range depending on the geometry and what's causing the obstruction — that's not catastrophic but at longer ranges where you're already working closer to the sensitivity floor it can matter.
The receiver sensitivity on the module is doing a lot of heavy lifting here, but it's not magic — Fresnel zone obstruction adds diffraction loss on top of free-space path loss and no front-end can compensate for physics. The practical thing I'd suggest is try to get 80%+ clearance if you're pushing toward those longer ranges, and pay attention to whether your obstruction is a hard edge (like a ridgeline) versus distributed (like tree canopy) since the diffraction characteristics are quite different.
What's your deployment scenario — fixed ground stations or one end is a drone?