This work is limited by a few metrics. The first is that the knowledge produced in this thesis was largely derived from analysis of existing 360 videos, and not through a trial-and-error approach where I would have filmed many 360 videos. More film theory and criticism than practical filmmaking. This affects the language and pragmatism of the content, and perhaps how "bleeding edge" the ideas could potentially be. It is not attuned to filmmakers but critics.
(Yet, filmmakers must wear the hat of critics in order to iterate.)
I believe there are advantages to this approach. One is it allowed me to analyze a wide variety of 360 films, not just the sorts of films I would be partial or capable of creating. Another is that it kept me focused on the analytical framework and toolset, and not on any one piece of technology, gear, equipment, or playback device.
This is, effectively, a practical solution to getting past the issue of production specifics, and the advice here is certainly broad.
This work remains device-agnostic and trusts that technology will improve, particularly in resolution. This means there is no discussion of minimum or maximum distances for recognizable faces, for example (see Naimark's work with the Google Jump camera system for non-device agnostic studies). The effort to avoid the influence of specific hardware or hardware limitations did not prevent this work from diving into the deeply pragmatic but did lend itself to be more suited to abstract and conceptual discussions.
This work did avoid an intense look into spatialized audio for purely practical, and not conceptual, reasons. If I could have spent more time studying this, I would have. Spatialized audio is hugely important, and the absence of its discussion here is a detriment.
This work also avoids a detailed look into the differences between stereoscopic and monoscopic 360° video, which would be beneficial, but is nowhere near as fundamental as spatialized audio. This is one of the elements of 360° video that I felt unqualified to write about without using the gear first hand. I would have liked to put a stereoscopic and a monoscopic camera side-by-side filmed the same thing, and observe the differences in effect on audiences. Without this, I could speculate and make educated guessed, but not anything with the degree of confidence that I talk about 360° video in other contexts.