The traditional method is long winded, time consuming etc etc.

Here is how we are doing it.
Run a reproducibility experiment over a minimum of 10 days. ( Same lab, same equipment, if possible different analysts, different reagents - Include as much variability as possible).
At the end of the experiment, determine the SD(Reproducibility). To get your measurement uncertainty @ 95% confidence, multiply the SD(R) by a coverage factor of 2. For 99% confidence, multiply by a coverage factor of 3.
Use this as your initial MU. Carry on with the Reproducibility experiment over time and review the MU periodically based on the information you get.
We use this approach for all our chemistry and micro analysis in the lab.
An alternative (to determine a starting point) is to use a certified reference material. Run a minimum of six replicates and follow the latest EURACHEM guide. (Drop me a mail if you want the calculations in EXCEL format)