If the equipment is "designed", not "tested" there is no need for calibration.
Not sure I understand the statement fully. Could you eloborate a bit on that please ?

a "designed" piece of equiment has all parametres verified in the design phase such as fan-out, fan-in, noise margins of 3.3V tresholds driving 5V, capacitive loading of high speed clocks.
A "tested" piece of equipment is put toghether without the above concerns, tested and "shown to work"

As an example related to your original post: If I were to make a tachometer that was to be reasonably precise I would not use the clock in the '51 I would use a well defined crystal oscillator. The internal clock in the '51 is not very well specified as to crystal load, basically it is specified as "it will oscillate". In extreme cases the need for calibration can be totally avoided by using a TCXO (temperature controlled crystal oscillator)

Erik