The simple equations of motion for waves on a string provide that
a general interference (as of two slightly-different frequencies)
will make an energy concentration that is textured at the square
of a wave at the difference frequency (which has energy lumps
at twice the difference frequency). While that is
amusing, it is also the principle that radios traditionally use in mixing a
received wave down to an intermediate frequency.
A more complete example of destructive interference is in the antireflection
coatings on camera lenses; the partial reflection from one layer may
completely destructively interfere with the partial reflection of another
layer, so that there is NO reflected light energy. The energy in the
incident light therefore is not lost to reflection, only the transmitted
light remains (and it is undiminished by any reflection). Thus, the energy
is again in a region which is NOT the region that corresponds to destructive
interference.
Unless a non-wave-equation term (such as absorption or gain) is present,
the wave interference phenomenon distributes energy in a variety of ways,
but does not lose it. It's not hiding, it's just... not particle kinetic
energy, so is not localized to a particle location.