move mag and SNR cuts to the end after linking?

[mschwamb]

Right now the mag and SNR cuts occur before the fading function so you can't reliably using the linking filter. Especially with the new filter to cut out observations that are far in apparent magnitude before emphemeris generation, we don't really need to make post-processing faster that I think was the original intent. Note this is not the 2 SNR cut.

Where these cuts are placed this is good for checking if a known orbit will ever get bright enough to be potentially observed in past LSST observations, but you might want to simulate a Col-OSSOS sample and you can't do that. We also might want to give options for the mag cut whether it's median, max, min of all observations or the discovery observations.

[mschwamb]

Also the mag limit cut is using the PSF mag, I think we want this to be the trailed source magnitude if we want to create a Col-OSSOS like sample - we should decide if want to change to trailed source magnitude