This also holds for the multivariate case:
DiracDelta[0] is not an "infinite" quantity:
DiracDelta can stay unevaluated for numeric arguments:
Products of distributions with coinciding singular support cannot be defined:
DiracDelta cannot be uniquely defined with complex arguments:
Numerical routines will typically miss the contributions from measures at single points:
Limit does not produce
DiracDelta as a limit of smooth functions:
Integrate never gives
DiracDelta as an integral of smooth functions: