Deconvolved size of 0, but multiple gaussians?

[jt2553]

PyBDSF deconvolves the source size with the synthesized beam. When the DC_Maj and DC_Min are both 0, then one can take that to mean that it is a point source. If either value is >0, it is a partially resolved source. A point source should have identical peak and total flux values (and island flux).
By this assumed metric, I have three point sources. All three have a significant difference between peak flux and total flux, and all three are made up of multiple Gaussians, as indicated by a S_Code of 'M'.
Are my assumptions above correct, and how can I reconcile the multiple Gaussian sources being point sources and having peak != total flux?