The question wether or not RS485 would work here is not the subject. It would without doubt. However, I indicated that it is not necessarily the best solution for all applications one could think of.

Taking into account that the field devices are only transmitting data, the option of a simple current loop, using the available supply lines, saves additional wires. Talking current loop, we talk low impedance, so the immunity to noise is high as well. As for that matter I don't see why RS485 would prevale over a current loop.

Then we have Eriks argument that current loops is something that died many years ago, merely because it has been used on teletypes, and the suggestion is made that it is only suitable for very low speeds. I disagree with that. Remember television back in the fifties? Modern TV receivers are not that much different from the ones that used to be made in those days, but we still use them. So I would rather call it 'proven' than 'old'. As for the baudrate: with a cable length of 100 meters (presumably a good quality cable) speeds of 2400 bps can be obtained without difficulty. If only a limited amount of data has to be transferred, what's the gain of using higher speeds? Nada.

So the only argument that remains standing here is that one can use a simple converter or interface on the PC side. Not the strongest, too, assuming that the poster of the question is quite capable of designing electronic circuits.

So what's the point here? What everbody obviously likes best is to open the drawer with 485's or similar and push one in your circuit. Easy, and especially: without thinking. qed!