This is not very different from collecting visual cues. You can notice a delivery van arriving. You can see the driver's face, same with passers-by. The biggest difference is that a camera needs to be more conspicuous, while a BT receiver can be invisible and undetectable. Much cheaper, too.

I have an ESP32 Cam in front of me right now. I think I paid maybe 8 bucks for it. If I wanted to, I could very easily hide the tiny camera in my front door, and use it to both collect bluetooth and wifi metadata (including MAC addresses) and correlate images/faces to MAC addresses when people pass by close enough so that I can identify them later from longer range wifi/ble detections.

(I actually do plan to install this at my front door, but aimed mainly to detect when a deliver/parcel in on my doorstep, and I don't (yet?) plan on sniffing bluetooth/wifi with it)