First, thank you for finding the original and follow-on posts.

Re; Guidelines which suggest content be in the post rather than the PDF. Here are some elements that are discussed in the PDF that suggest it's not fake. Good or bad, they are elements to think about:

1. The PDF discusses CGI requirements circa 2011 which are required if this is CGI. CGI is not dismissed, rather, deprecated. CGI is possible, but for the time of recording, not likely.

a. optical features consistent of imaging through dual pane safety glass are present (page 3)​

b. refraction of the object when NOT in the field of view is present (pages 5.6.7)​

cl. During recession, the final ancillary flash occurs partially obscured by a tree with pixels indicating partial light transmission through greenery. (page 8)​

​

​

Assuming non CGI origin, window reflections of interior lights are possible, but not consistent with the 3-d reconstruction that follows. i.e. the vehicle is not large enough to show a reflection at an estimated 100 foot distance.

The 3-d analysis is the next section. Because the vehicle from which the recording originates is in motion it is possible to create a 3d stereo-image using the Pulfrich effect, whereby one eye views the current frame while the other eye views the previous frame.

This permits an approximate measurement of the distance of the object from the camera. When the object reaches the edge of the road, it is at a distance comparable to a previous view of a road sign type which permits an approximate measurement of the object diameter as 1 to 1.5 feet. This further leads to a closest approach distance of approximately 10 feet assuming the Android camera is properly identified.

The PDF looks at possible real-world equivalents and faults ball lightning and laser induced breakdown as candidates.

So as the analyst here, a decade later, I have no idea what it was, and if faked, I have no idea how it was faked. Hence, I've posted here in hopes someone can give me better answers than I've come up with.

​