Nope you didn't miss it. We currently don't have a way to take the outputs from the Point Estimator and load them into the datalink. It is on our to-do list, but for now it is really just meant to be used as a guide. I am pretty sure you can open the KML in the DJI datalink software, but I could be wrong. When we do make it spit out the correct output, it will likely only work with the PC version and not the iPad.
All cameras, including the GoPro, get set to take periodic pictures. The Point Estimator will give you a value which is needed to get proper overlap. You need to take images AT LEAST that fast.
GoPros don't geotag their images so they can't be used in the easiest EXIF based workflow we offer. The next best thing, and probably more accurate, is to use the basemap georeferencing workflow. This will allow you to associate points within your images with points on an existing basemap (Google Maps or something similar). Another option is to use a GPS recorder and use a tag inserter like GeoSetter.
It is important to use the narrowest field of view (lowest resolution) setting on your GoPro when planning your overlap and collecting images. The field of view of the GoPro is too wide at its full resolution to be able to make accurate maps with. The center portion which is used in the lower resolution modes is much better. Even better than using the low resolution setting is to get a longer focal length lens like the RageCam lenses or the ones we install and calibrate at DronesMadeEasy.com.
I hope that helps.