I had a nice chat with my chemistry teacher today. He did not seem to think that conductivity would give an accurate ppm reading because we don't know exactly what is in the water. It could be different ions, and salts, and minerals, we just don't know. All the different things would have different amounts of resistance, and some would not be conductive at all, these would go unaccounted for. He recommended boiling down the water and measure how much stuff is left over.
If we were looking for total ppm boiling would really be the only way to get a true answer, but remember that a digital TDS meter measures conductivity. We aren't looking for things that can't be dissolved in the water though. You can have as much SiO2 (silica) floating in the water as you want and it's not going to make any difference to the plant because it's inert. We're mostly concerned about things that are ionically bonded and easily dissociated by water, namely salts. Since the ions are easily dissociated by water they're going to make the water more conductive, although the different salts are going to have different conductivities (refer to the table here http://www.appslabs.com.au/salinity.htm ). Since they have different conductivities we've kinda been running with a worst case (NaCl is a factor of 0.51, we've been using 0.70).