For in-camera stacking, it's basically trial and error. Moreover, in-camera stacking requires you to set the initial focus point about a third of the way to half-way into the depth of the subject because it moves the focus point towards the camera for a few frames and then away from the camera for the remaining frames. Once you figure out the step size for the number of frames for that distance and that aperture, it can work great if you are lucky. If you are off in setting the initial focus point or you have the wrong step size, you may not get everything in focus. If you are shooting at 1:1, remember that the focus point will try to back up, so make sure that is possible when you set the initial focus point or it won't be able to focus on the nearer part of the subject. Moreover, the camera's stacking algorithm for aligning and compositing frames is pickier than Helicon or Photoshop's. In-camera stacking saves all of the individual frames as well as the stacked image (which will be a JPEG only). Of course, you always have the option to stack the individual frames in post if the camera can't handle them, and your post software might be more forgiving than the in-camera software.
In contrast, the "bracketing" feature lets you set the initial focus point on the nearest part of the subject and then it moves the focus point only away from the camera. This makes a lot more sense to me. In addition, bracketing will allow you to shoot up to 999 frames, but you have to stack them in post. Setting the step size is the key here, because if you shoot a larger number of frames than needed you can simply discard those that are past the farthest part of the subject. This also requires some trial and error to avoid OOF bands in the final stack, but it is more straightforward IMHO.
