Broadly speaking I would suggest the ARM should adapt apropriately in consideration of what you've mentioned in one cluster (maybe somebody else knows differently). All single clusters I've seen worked it out just fine even though IAPs were dense in some spaces, light in others where needs dictated.
However, by simple virtue of the fact you've queried it, I'd recommend you have a quick look at how the ARM has converged once it's in flight in a typical day, and check it "looks right" with your RF head on.