Recognition of the limits of human perception is a necessary component of analysis. That is something you dismiss at your peril when you cling to your conclusions of the temperature of the falling components in the video. You haven't accounted for the inaccuracies in question, again not the least of which was, as mentioned above, the color calibration of the monitor the image is viewed on, as well as the accuracy of the camera in question. And this is all before the human inaccuracies that is the point of this response. Note the information at the link you yourself provided earlier.
Tables comparing temperature and color appeared at least as early as 1836 (Pouillet). The one below shows three attempts at correlating temperature and color. The verbal descriptions given by Howe2 and White and Taylor3 have been omitted and their temperatures placed with the verbal description in the Holcomb Steel data that was closest to theirs. The variation demonstrates how unreliable this method is even in the hands of careful observers.
http://www.sizes.com/materls/colors_of_heated_metals.htm
As an example of the table at that site, note the temperature differences between the Halcomb Steel descriptions and the White & Taylor ones: Orange-yellow is 1300 degrees C in one as opposed to 941 in another. And that's with
careful observation of material in person, not 3rd party observation of a recording, with all the additional error that is added by the recording device.
Keep in mind that when you discuss quantitative conclusions - such as the temperature of phenomena observed at the twin towers - qualitative assessments lack the precision necessary to draw such conclusions. The charts at the link above demonstrate this. Proper application of empiricism understands the difference between qualitative and quantitative assessments, and proper analysis of phenomena understand the errors that can occur in both equipment used and human observation. And you'd hope that the error in the equipment is measured or at least accounted for, not dismissed by increasing the numerical range of the conclusion drawn. You are not keeping any of this in mind, not when you defend your conclusions as being as based in perception as anyone elses. Yes, your assessments are based in perception, but your failure to take properly into account the sources of error and limits of such perceptions are what render your conclusions suspect.