hard-coded values in get_conversion_factor

[PBarmby]

The hard-coded number 2.3504 * 10**(-5) should be compute-able with the units package: it's just the conversion between MJy/steradians and Jy/arcsec.

I was wonderng if it would be better to store some of the other numbers in some kind of 'calibration database' so that the code doesn't need to be updated -- but I think astropy discourages this?

[sophiathl]

i like the idea of a calibration database - I haven't seen something like prohibiting this on astropy specifics, but we will check to make sure.

[sophiathl]

Calibration database still to be explored/implemented