I did show the numbers which you even quoted and which are based on looking at failures of Windows 10 upgrades reported in threads just like this one, as I wrote. You may be even right hinting at failure rates that may be way larger than 0.01%, take that number then as a label for 'way too large' ('way too large' for work computers, that is. Both work and hobby computers are affected, but the failure rates may be adequate for hobby computers)
Not sure this is a hobby site, there are certainly small business owners on this site that maintain their own work computers (or have their grandpa on this site that does the Windows 10 upgrades for them
)
Say that upgrading a work computer to Windows 10 fails and the financial loss incurred is $1K per day. Say it takes 2 days to get it working again. It would be cheaper to get a new work computer in the first place, or not upgrading at all and holding out with an old OS as long as possible. The latter is very common, even for large businesses and government agencies. And no, these people are not all dummies, they have things to consider that hobbyists do not have to consider.
These considerations apply to all kinds of work computers: Windows, Apple, Linux, etc. As I wrote further above, in the past these considerations made me buy new work computers rather than upgrading old work computers (I did upgrade my hobby computers)
The possible game changer is SaaS (Software as a Service) that Microsoft is touting for Windows 10. But it would only be an actual game changer if upgrades don't fail (or have a failure rate of near zero). With the number of failures reported in threads like this, a professional running a home office, a dentist, etc. would certainly not upgrade. They would just buy a new work computer if the old one stops working, or do a scheduled replacement every 3 years or so, it is cheaper that way.