The camera integration currently allows displaying on the dashboard the image from a camera publishing an RTSP stream, and activating the video stream. You can also send a camera image to Telegram from a scene. But we could do more 
When a camera is ONVIF-compatible (this is, for example, the case for TP-Link Tapo cameras), it is technically possible via this standardized protocol to control the camera. See explanations here.
I think it would be useful in Gladys to be able to access the following functions from a scene or the dashboard:
- enable/disable a camera
- change the camera’s « horizontal » orientation (pan)
- change the camera’s « vertical » orientation (tilt)
- change the camera’s zoom (zoom)
- (and if possible) position the camera to one of its presets (pan+tilt+zoom)
- manage the camera’s motion detection (to use it as a scene trigger)
- Broadcast text via ‹ text-to-speech › on a camera (by making a camera selectable in the action « speak on a speaker »).
Some use cases:
- when I leave the house, enable the cameras (and vice versa when I return)
- every evening, when I’m away, send me on Telegram a photo from my camera by successively looking at different angles of the room
- in case of intrusion, broadcast a deterrent message on my camera
- …
From what I understand of the ONVIF standard, it is the « Profile S » that should be taken into account. The ONVIF profiles are described here.
I don’t know if this can help development, but there is a Home Assistant plug-in available on GitHub that handles the ONVIF protocol for Tapo cameras: here