You said, "it's a good theory because there is no evidence of where the chips tested by Jones originated."
If there is no explanation of where the chips originated then you think that's a reason to say there's nothing suspicious? I'm not getting your logic.
Regarding different testing conditions, that is quite a stretch. That the same material would produce such a conspicuous and energetic reaction consistently between 415 °C and 435 °C in one set of tests but then not combust at all to temperatures beyond 800 °C in another set of tests, where both tests are a lab attempt at doing nothing more than heating the materials, seems pretty unlikely to me.
NIST doesn't tell us the details of their tests, but the fact that they were able to control the temperatures and maintain them over extended times implies that they put their samples into some type of chamber designed for that purpose. I'm doubtful there is any meaningful difference between the two sets of tests, or any that could explain such a dramatic difference in the results anyway.
Edit to add: NIST tells us that "above approximately 800 °C, ... a thick scale formed and spalled off of the steel, carrying away the paint." So your idea that the paint being attached to steel would explain why it didn't combust does not work. Even when the paint separated from the steel, it still did not combust at temperatures above 800 °C. Regarding the possibility that the chips could be common primer paint rather than the more specialized ceramic coating, Harrit et al addressed that question in their article (emphasis is mine):
7. Could the Red Chip Material be Ordinary Paint?
We measured the resistivity of the red material (with very
little gray adhering to one side) using a Fluke 8842A multimeter
in order to compare with ordinary paints, using the
formula:
Specific resistivity = RA / L
where R = resistance (ohms); A = cross-sectional area (m2); L
= thickness (m).
Given the small size of the red chip, about 0.5 mm x 0.5
mm, we used two probes and obtained a rough value of approximately
10 ohm-m. This is several orders of magnitude
less than paint coatings we found tabulated which are typically
over 1010 ohm-m <31>.
Another test, described above, involved subjection of red
chips to methyl ethyl ketone solvent for tens of hours, with
agitation. The red material did swell but did not dissolve, and
a hard silicon-rich matrix remained after this procedure. On
the other hand, paint samples in the same exposure to MEK
solvent became limp and showed significant dissolution, as
expected since MEK is a paint solvent.
Further, we have shown that the red material contains
both elemental aluminum and iron oxide, the ingredients of
thermite, in interesting configuration and intimate mixing in
the surviving chips (see Results, section 1). The species are
small (e.g., the iron oxide grains are roughly 100 nm across)
in a matrix including silicon and carbon, suggesting a superthermite
composite. Red chips when ignited produce very
high temperatures even now, several years after the 9/11
tragedy, as shown by the bright flash observed and the pro-
duction of molten iron-rich spheres (see photomicrographs in
Fig. (20) above). Correspondingly, the DSC tests demonstrate
the release of high enthalpy, actually exceeding that of
pure thermite. Furthermore, the energy is released over a
short period of time, shown by the narrowness of the peak in
Fig. (29). The post-DSC-test residue contains microspheres
in which the iron exceeds the oxygen content, implying that
at least some of the iron oxide has been reduced in the reaction.
If a paint were devised that incorporated these very
energetic materials, it would be highly dangerous when dry
and most unlikely to receive regulatory approval for building
use. To merit consideration, any assertion that a prosaic substance
such as paint could match the characteristics we have
described would have to be accompanied by empirical demonstration
using a sample of the proposed material, including
SEM/XEDS and DSC analyses.