@Mary Wilson yeah RSSI ranging is definitely doable for that kind of coarse bucketing — "within 500m vs around 2km" is a much more forgiving target than trying to get meter-level precision, so you're working in the right regime.
That said, raw RSSI on any RF link jumps around more than you'd expect, and the MultiNav Pro+ is no exception. The external amplifiers and the receiver sensitivity that give it that long-range performance also mean you're picking up a lot of signal variation from multipath, orientation changes, and environmental stuff — trees, terrain, even humidity. A single reading is pretty much useless for distance estimation, but if you median-filter or exponentially smooth over a rolling window of maybe 10–20 samples, the values stabilize enough to reliably distinguish the kind of coarse distance bands you're describing. The difference in RSSI between 500m and 2km on these modules is large enough (we're talking 20+ dB in clear conditions) that you've got plenty of headroom even with noisy readings.
One practical thing worth doing: take a few reference measurements in your actual deployment environment — RSSI at 200m, 500m, 1km, 2km — and build your thresholds from those empirical values rather than relying purely on the free-space path loss model. Real-world terrain changes the curve enough that hardcoded theoretical thresholds will let you down. Once you've got those calibration points, a simple RSSI bucket check with some hysteresis around the threshold should work reliably for what you're building.
