I really think just having raw pulse output from flow sensors can be useful by itself. In my case, I have 2 flow sensors installed on my irrigation lines. One is at beginning of mainline by shut-off valve, and a second one is at the end run (downstream of flow) of a valve which only waters a small garden of tomato plants. In the latter case, water use is minimal but I like to check that its programmed runtimes are operating normally.
Mainline (2-wire) flow meter is built to output 1 pulse to 0.1 cubic ft. or 0.748 gallons. There are at least 8 valves which would pull water from this mainline. So monitoring overall water use is the only function of that flow meter.
OTOH my second valve has a flow sensor(3-wire) which monitors a couple of micro-irrigation spray nozzles. It is programmed to water each day, 4xday, but for 1 minute only. I see on graph of pulse transmitter #118134 that these watering events are taking place because I see four peaks of flow with graph of pulse data. That is on page 1 of link below:
http://tinyurl.com/flow-sensor-graphs
If you scroll further down the doc to second page, I have a graph of (wireless) pulse transmitter #301214, I think it’s evident that it’s impossible to detect such a small amount of water use from the second valve by looking the ‘big picture’ graph of mainline flow data. So in this case, scaling pulses to gallons (or milliliters) is totally unnecessary and mostly an exercise in futility.