I've read all of the comments on the HotRodders forum and feel that there is something missing. It should be noted that I've never seen any other tests of this type so anyone can comment from opinion but they should either do a test or quote from another test result that they have read. Opinions are sometimes well intended but have little or no value unless they have some experimental data to back them up. Serge may not have covered ALL the variables but he did a decent job of documenting the variables that he did cover.
As far as allowing enough time for the primer to cure properly it was stated that it should be left for a week. This may be a good "rule of thumb" but it can vary with temperature, application and the materials that were used. How many pro shops do you know that would wait a week before applying filler to a job? Not many I'll bet.
I sat on the board of a local body shop association when the epoxy primer recommendation was implemented by certain paint manufacturers and I found that no shop owner I questioned used primer under filler as a matter of practice.
In our shop we use primer under filler when the filler is applied over a seam located low on the vehicle. We do this because the seam could allow moisture to reach the face of the metal and the primer could help shield the metal from the moisture. In 39 years of automobile restoration we've never experienced any failure of filler bonding directly to the metal as long as the metal was prepared properly and the filler was applied properly.
One other note: As far as I know there was no metal conditioner used prior to the epoxy primer being applied. This would have introduced another variable that may have helped or hurt the adhesion depending on how it was used.