I stand corrected. I am a victim of "I saw it on the internet so it must be true".
However, I avoid using Phillips screws because the do cam-out easily, although probably because the driver has gotten out of line with the screw.
The caveat of using a JIS driver on JIS screws still holds true.
And personally, I find that a JIS driver works better in a Phillips screw than a Phillips driver.
Ironically, there is a "camming action" described in the original, 1936 patent, but this is an action that is said to clear the screw of any contaminants, not the one that removes the driver from the screw:
*************************************************************************
"Another object of the invention is the particular angular formation of the walls of the recess in the screw with respect to the angular formation of the working end or bit of the driver to establish a wedging engagement between the two when united. This same angular formation of both elements is especially designed to also create what might be termed a camming action during the approach of these angular faces toward one another with respect to any substances which might have become lodged within the recess of the screw. It has been found by experiment that a downward thrust of the bit into the recess will instantly dislodge any substance within the recess by causing it to move upwardly and outwardly over the walls of the recess."
*************************************************************************
It seems possible that this part of the original patent description was confused in a quick reading with the "camming out" of the driver from the screw head, so leading to the myth that the latter was intentional. Who knows.
It also seems possible that this angular arrangement intended to clear substances from the recess was also conducive to camming the driver out of the recess, once there was wear on the driver or screw head.
The idea of using a JIS driver in a Phillips head sounds reasonable.
Moto