Seems that some people are confused about EM wave
interference. Here's a quote from "Optics", by
Hecht:

"Optical interference corresponds to the interaction
of two or more lightwaves yielding a resultant irradiance
that deviates from the sum of the component irradiances."

The units of irradiance are joules/sec/unit-area which
are the same units as the Poynting (power flow) vector.
So putting Hecht's interference statement into words
familiar to RF engineers, we can say:

RF wave interference corresponds to the interaction
of two or more RF waves yielding a total Poynting
vector that deviates from the sum of the component
Poynting vectors.

It follows that if the total Poynting vector does
not deviate from the sum of the component Poynting
vectors, then there is no interference, by definition.

Here's the kicker to this discussion of interference.
Assuming the following wave components are moving in
the same path in the same direction:

1. Superposing two mutually incoherent EM waves of equal
magnitudes results in a maximum possible intensity that
is the sum of the maximum intensities of the two waves,
i.e. two times the maximum intensity of a single wave.

2. Superposing two coherent EM waves of equal magnitudes
results in a maximum possible intensity that is greater
than the sum of the maximum intensities of the two waves,
i.e. four times the maximum intensity of a single wave.

In a nutshell, that is the difference between waves that
interfere and waves that do not interfere. Light waves
that interfere can be twice as bright as light waves
that do not interfere. The constructive interference
term can double the intensity of the sum of the
intensities of the component waves. The extra energy
is supplied from areas of destructive interference
(or by a local source).
--
73, Cecil http://www.w5dxp.com