No, I think I see how it works, and why it doesn't. He notates negative exponents from 1 to 9 by subtracting the exponent from 10, from 10 to 99 by subtracting it from 100, from 100 to 990 by subtracting it from 1000, ... The result is that you can multiply by simply adding exponents and then dropping a leading 1; except that sometimes you have to multiply by adding exponents and not dropping the leading 1. There will be a rule determining which you have to do, but that adds an operation that isn't necessary with traditional scientific notation, so at best his notation adds one extra step. This is a result of his idea that, for a negative exponent, any number of leading 9s may be discarded; unfortunately, that idea simply doesn't work infallibly.
Dave