Remarks on quadrilateral finite elements for a fluid-structure eigenproblem