I have read in my textbook that the motivation behind the existence of FIR filter is "In particular, FIR filters are used for their ease of implementation and stability."
The stability issue is quite easy to explain from the transfer function along with using certain stability criterion.
But I don't understand why they are easy to implement. Compared with what? With IIR filters?
They are easy to implement because they require only 3 elements: storage memory for accumulator value, and multiplicate and sum units.
IIR filters also req the same units but they stability is uncertain and accumulator length is much more complicated to calculate so that it will not overflow. Not mantioned that in real hardware loopback is hard to code correctly.