@Patricia Moore good question, and you're right to flag it — the data rate vs. sensitivity tradeoff is real and it absolutely matters at these distances.
The 50km spec is the drone-to-drone figure, meaning both modules are airborne and you've eliminated ground reflections and multipath interference. That's the best-case geometry, which is why the number looks so dramatic compared to ground-level deployments (which are rated at 5km). The range performance here comes primarily from the external amplifier stage and the receiver sensitivity optimization Moamen built into the hardware, not just from PHY selection — so it's not a straightforward "just drop to 125Kbps coded PHY and you get the max range" situation like you'd see on a bare Nordic chip.
For your open field telemetry use case, you're in a good spot because you're describing conditions close to that best-case geometry. That said, if you want to dig into exactly which PHY mode those range specs were validated at, I'd post the question directly to Moamen in the Dev Hub or check the Resources section — the technical documentation should have the test conditions, and if it doesn't, that's exactly the kind of spec clarification worth asking for before you build your deployment around it.
What's your actual data throughput requirement? If you're streaming sensor packets and not video, you may not even need 2Mbps and could run at a lower rate to buy yourself extra link margin, which would help on the reliability side even if raw range isn't your bottleneck.
