Having been doing this for over 5 years, it’s unfortunate Youtube has made things harder. Being a local Public Access TV producer, I first saw the viability of 360 video from the FIA’s Formula E racing (car) series, and quickly adopted their tested and incorporated technology. It involved an Allie camera (initially conceived as a real-time “VR-steerable” security camera) which stitched (putting together the video from 2 or more cameras/lens) in real time! Better, uploading to Youtube was just a single button click (as they had an agreement with Youtube). Unfortunately, things have changed, and now Google/Youtube demands everyone use an encoder (what happened to clicking one button?)!
So, with my experience/expertise, the biggest thing I can tell anyone embarking on 360, or 360x360 video, other than watch out for time and processor power required to stitch certain cameras’ video, is that it’s different: you no longer point and shoot. Instead, it’s easier; all you do is hold (or set up) your camera, and it catches everything (including you if you’re holding it). About the biggest problem is making sure you yourself aren’t blocking some of the action you want seen by the sphere that’s being videoed!