I've tacked down the failed test here to a regression in SymbolicUtils v3.13.0 released Jan 31st
on SymbolicUtils v3.12.0
julia> qsimplify(dagger(DisplaceOp(im)) * Create * DisplaceOp(im))
((0 - 1im)π+aβ )
on SymbolicUtils v3.13.0
julia> qsimplify(dagger(DisplaceOp(im)) * Create * DisplaceOp(im))
D(im)β aβ D(im)
# what's really weird is that the other similar rule is being applied fine
julia> qsimplify(dagger(DisplaceOp(im)) * Destroy * DisplaceOp(im))
((0 + 1im)π+a)