To compare these measurements against a generic DMM measurements it would be arguable, to say the least. The comparison against bench/professional meters has already been done by the developers and other supporters, and that's way they have come with several recommendations on this thread and the user manuals, and some improvements on the firmware.
To make a summary of those, IIRC, the accuracy of a component tester depends on a)the Voltage regulator precision (5%, 1%, etc.), b)the Voltage reference presicion (which should have 10x times the precision of the Voltage regulator, or be removed, and configured accordingly in the firmware), the frequency of the crystal (deppending on what you're measuring), and other factors such as whether the device has been calibrated the way is being used, and even whether is being powered from a clean steady DC voltage, or something with more or less ripple.
Bottomline, the accuracy/repeatability of the measurements can vary from one device to another.