I've also never heard of this. However, the math is based on using a scale factor. If you can figure out a suitable base-ten scale factor for your input, then you can do the math in integers with all of the significant digits, (as you've shown) and present the output in decimal format.
