A ug 2 00 9 On Convergence Rates Equivalency and Sampling Strategies in a Functional Deconvolution Model