Yeah, good luck with that.

The CLEAN algorithm relies on the PSF being identical to the analytic PSF for every point in the image. For the central source with its partially inflated sidelobes it won't be successful, and will leave residual PSF-like emission around it.

You can see the result below for both the unperturbed sky and the double flare sky. Note again the sigma values in the Purr FITS renders. Awesome as Purr is it often doesn't do the most photogenic job of making PNG out of a FITS file, so I've attached two images (on the same colour scale!) below, exported directly from Tigger.

This residual PSF emission around sources is usually a good sign that something is amiss (unless of course you are interested in astrophysical transients -- you don't want to calibrate out your Nature paper).

Can you think of a calibration approach that would take care of this problem?

Note that the CASA imager / lwimager will only perform a clean operation on the CORRECTED column of a Measurement Set, so I had to play a quick game of "Musical Columns" to get the images below.