glfp, Image size (resolution), Image quality/properties and frame rate are what determines the bit rate. Adjusting the bit rate is not a quality setting in its own right, more a means of control over network throughput. So looking at the HikVision table, with a 2MP camera at a resolution of 1080P running at the highest quality setting at 15 fps then if you allow for more than 5120 (for example 8192) then you will achieve the optimum recorded image. However, if you set it to limit at 4096 then the camera will adjust resolution and quality in order to meet this limit. So the bit rate adjustment is more like a throttle, you can leave it wide open for best performance or close it off for best economy. If you are worried about recorded file size then FPS probably has the biggest impact. H264+ compression can have an advantage as it is a smart codec that applies different levels of compression to the image. As I understand it static areas of the image are compressed more than dynamic areas, so the walls of a corridor might be highly compressed whilst people moving down the centre of the corridor are less so (I am not big on algorithms so there is probably a better way to explain this).
As an experiment, maybe you could run the camera in VBR mode and try different image quality and FPS settings and see what that affect that has on your camera output. Based on that, set a Maximum bit rate value that you are happy with and that meets your storage requirements. I am still learning myself (and enjoying it) but I hope that I have helped you in some way. There is also plenty of good help on the web if you dig around.