Should We Believe Atmospheric Temperatures Measured by Entry Accelerometers Traveling at "Slow" Near-Sonic Speeds?

Mars Pathfinder's Accelerometer instrument measured an unexpected and large temperature inversion between 10 and 20 kilometer altitude. Other instruments have failed to detect similar temperature inversions. I test whether this inversion is real or not by examining what changes have to be made to the assumptions in the accelerometer data processing to obtain a more "expected" temperature profile. Changes in derived temperature of up to 30K, or 15%, are necessary, which correspond to changes in derived density of up to 25% and changes in derived pressure of up to 10%. If the drag coefficient is changed to satisfy this, then instead of decreasing from 1.6 to 1.4 from 20 kilometers to 10 kilometers, the drag coefficient must increase from 1.6 to 1.8 instead. If winds are invoked, then speeds of 60 meters per second are necessary, four times greater than those predicted. Refinements to the equation of hydrostatic equilibrium modify the temperature profile by an order of magnitude less than the desired amount. Unrealistically large instrument drifts of 0.5-1.0 meters per square second are needed to adjust the temperature profile as desired. However, rotational contributions to the accelerations may have the necessary magnitude and direction to make this correction. Determining whether this hypothesis is true will require further study of the rigid body equations of motion, with detailed knowledge of the positions of all six accelerometers. The paradox concerning this inversion is not yet resolved. It is important to resolve it because the paradox has some startling implications. At one extreme, are temperature profiles derived from accelerometers inherently inaccurate by 20K or more? At the other extreme, are RS temperature profiles inaccurate by this same amount?