I'm well aware of the parallels to Microsoft, another conservative company. Just because it happened once, doesn't mean it will happen again; by the same token, just because it happened to another company doesn't mean Canon will learn from that lesson. Usually, changes like this don't happen overnight. Obviously, Microsoft did not respond to those changes appropriately for their business. It remains to be seen how Canon will respond to a declining marketshare, but of course, first that marketshare has to actually decline.
That's an interesting comparison because I do know one thing or two about Microsoft because I worked as a programmer for a MS Gold Certified partner some time, gaining insight how that company works - and it's not like Canon.
Microsoft has missed main developments again and again like the Internet or the mobile market - *but* they were always able to turn around 180 degrees and renew themselves like with Internet Explorer and now semi-abandoning WPF/.NET/Silverlight and multiple other technologies that they promoted for years a the future way. That always left a lot of victims in the M$ crowd who weren't able to turn around as fast as their tech base did, but it saved Microsoft until now.
Canon seems to be conservative because they lack imagination, while Microsoft is conservative because they use their "embrace, extend and extinguish" strategy (at least up to now) which needs a lot of time to work, i.e. pinning people to M$ tech.
I'm keen to see if Canon will be able to turn around from their strong product differentiation (high price or low features/quality) strategy in case they face a problem down the road, but on the other hand maybe they just did with the eos-m and we didn't notice until mirrorless reaches the old school dlsr segment with the cash cow customer base (that's us).