Thanks for the reply, however I'm a bit surprised.
I'm not talking about writing major codes for interaction and walking. The current interaction method uses the camera to rotate and find the object to interact with, if the object is not visible. You could easily write your own interaction method setting operateCamera to false and setting an arbitrary miniMapDistanceThreshold to make it walk using the main screen. I think (hope) you know what I'm referring to. By this, the fixed camera setup would make the script less bot-like because of the unnecessary clicks on minimap, while rotating the camera. This is a BIG detectability imo.
Also, as I'm sure you know, upgrading the wheel to make it more efficient is often a good idea. Imagine if cars were driving with wheels made of stone rather than rubber.