The following experiments must be conducted:
- Compare results when the background is fixed to be the residual map from deconvolution and when is set to be a multiplicative scalar of it (e.g. an additional parameter to me minimised)
- Compare results with a simple flat-sky model
- Compare results when a bkg estimator is used as a model of the background