a) We don't actually know that it worked for GE. There were a lot of changes under Welch, so there were a lot of simultaneously changing variables. I would contend that the "10%" policy had smaller potential than lot of the other changes. Note that one of the changes was changing from a company that was primarily an "old guard" manufacturing business to one whose net income is nearly 50% from its financial subsidiary (http://en.wikipedia.org/wiki/GE_Capital) - before the recession, it was substantially more than 50%. That is a huge change.

b) GE under Welch was an old, established, stagnant company. My personal theory (pure speculation) is that GE had built up a lot of inefficient processes: whenever something bad happens, regardless of the probability of it recurring, the natural response of (especially large) corporations is to write another procedure to prevent it from recurring. GE also likely accreated lots of ineffective employees over the years (to write and follow the ineffective processes ;-).

As a result, GE under Welch was able to cut a lot of costs out of their manufacturing processes by re-evaluating the cost vs. effectiveness vs. benefits of them. All those cut costs would flow to their bottom line, at least for a while.

Note that GE's stock has been pretty flat (struggling, even) for the last several years. According to my speculation above, this would be because they squeezed all the excess costs out of their manufacturing and their financial division took it on the chin with the downturn, leaving them sucking wind. So...

c) Jack Welch's changes happened in a different era and were in response to a different situation. A solution that works for a totally different company than Microsoft in a totally different era isn't likely to be effective when applied blindly to e.g. Microsoft.