Interaction is a complex question, which includes identifying
what sort of interaction you are concerned about, and specific
details of the antennas and installation.
Many hams have successfully run antennas in close proximity,
even interlaced. The pattern might be distorted somewhat,
depending on the details. Keeping the antennas at right angles
to each other will often help to reduce interaction.
In this case, if the dipole is mounted above the delta loop on
the same mast, the interaction will be greatest when the dipole
is in the same plane as the delta loop, and less when they are
at right angles. Meanwhile, interaction will typically be less for
a vertically polarized loop, but the actual polarization response
of the loop on 20m will depend on the feedline connected to it
when the other antenna is in use. (While the loop will be resonant
on 20m whether the feedpoint is open or shorted, the current
distribution around the loop will be different. Any reactance
added by the coax stub will affect that as well.)
If you were using a beam antenna with a sharp rear lobe in order
to null out signals, the interaction would be more important, as
it could shift the null. If you're just wanting a "generally
omnidirectional" radiation pattern, then a couple dB variation
probably won't make a significant difference in your operation.
The coax to the 20m dipole may affect the pattern of the delta
loop, depending how it is placed. If it runs straight down the
mast in the center of the loop it probably won't be an issue.
So there are lots of unknowns. I'd say your chances are pretty
good that it will work "well enough" with 5m spacing, so you can
try it and see. Assuming you have the antennas installed using
pulleys and halyards (at least for the top of the loop), then you
can raise and lower the loop while listening on the dipole to
get an idea of the interaction.