When I read definitions like that I get the distinct feeling that I don't fully understand the terms being used. So my layperson's way of thinking about entrance pupil is to imagine that the entrance pupil is the size and location where a single lens would be found if the complex multi-element camera lens could be replaced by a single lens element. That may not be technically correct, but I think it is close enough to serve our purposes.
Except with perhaps extremely simple lens designs, the entrance pupil is not the front lens element. It is also not the iris diaphram itself. It is usually (perhaps always?) located somewhere between the front lens element and the rear lens element. The following two links have diagrams of some different lenses and show the location of the entrance pupil. As you can see, the entrance pupil is not always located in the same place - though they do tend to be toward the center of the lens.
The entrance pupil can be located sufficiently for our purposes by trial and error if you have an eyepiece with sufficient eye relief. Zoom out the camera until some vignetting is visible. Then move the camera closer and farther away from the eyepiece until you find the location where the somewhat vignetted view matches the view seen through the eyepiece with your eye. In other words, where the items at the edge of view through the telescope are at the edge of view where the camera vignetting is seen.
You can make a good estimate of whether the camera lens' entrance pupil location is favorable for afocal coupling (close to the front of the lens) or not by observing the how the shutter/iris lens assembly moves when the camera is zooms through its range of focal lengths. The cameras that offer the greatest range have the iris/shutter very far forward at full wide, it then drops back some at around midzoom but begins moving forward again as the camera comes to a full zoom. Other cameras exhibit the back and forth motion but in the opposite way so that the shutter/iris is very far forward during mid-zoom. Cameras that have iris/shutter mechanisms that simply fall backwards (and sometimes dramatically so) as the camera zoom in increased area consistently require extraordinary eye relief in order to get unvignetted views. This is typical but not exclusive to zooms with very large focal length ranges (5x to 12x). There is, theoretically at least, another way. That is to locate what panorma photographers often mistakenly call the camera's "nodal point". In fact, this is the entrance pupil pupil of the lens and not the "nodal point". Even if the term is wrong, the technique is apparently correct. Here are three pages that discuss methods of locating the entrance pupil.
I have not tried these methods. It is also possible that these methods, even if they accurately locate the nodal point may not be sufficient. I'm not at all sure whether the camera lens itself will effectively shorten the eye relief of an eyepiece. For instance, it is quite possible that the entrance pupil of a CP995 is actually located 15mm from the front lens window but that the camera's lens elements cause the rays of light from the eyepiece to converge at closer distance. Consequently, 20mm of eye relief is needed even though the entrance pupil is only located 15mm away, So perhaps the camera lens' entrance pupil location by itself might not be sufficient. I will update this section when and if I learn more.