Take a look at all the following molecules: $\ce{SO3^{2-}}$, $\ce{SO2}$, $\ce{SO4^2-}$, $\ce{PF5}$ and many more you can add. In all these there is a controversy of one of these two types:
Is formal charge more important than octet rule - as to justify hypervalency- on determining Lewis structure? As a rule I'd say the answer is no. For example in $\ce{SO2}$,$\ce{SO3}$, $\ce{SO4^2-}$, Do calculations show it?
What is the central atom hybridization? For example, in $\ce{PF5}$ (see here). Does it involve d-orbitals?
In any case, both previous statements are related to the d-orbitals energy.
There is -at least- one straightforward way to answer the question:
- Calculate* the Molecular Orbitals and see which of previous options fit better (the use of d-orbitals or the non use of them). *sure some can be found in some papers.
So the question is:
Are there theoretical studies of many concrete examples as to remark that d orbitals don't play any relevant role in covalent bonding?