Yup, this is likely a phonological restriction in addition to a syntactic one, though it’s worth noting that the copula (the “be” verb) shows a lot of idiosyncratic behavior in different contexts in different dialects of English.
It seems that this pattern may have something to do with stress assignment within a predicate, but I’m not sure what the conditioning environment is at first glance. Any English phonologists here who can shed some more light on this?