Considering the price of a cheap desktop these days, it is probably a good idea anyway.
Having to replace the PC every 8-10 years is not that bad.
My main home computer is a 7 years old desktop which runs Win7 just fine, so I assume any PC not capable of running Win7/8 is close to 10 years old?

Then again, with an updated antivirus and being a little careful with the browsing, the risk is not really that high anyway.

Considering the cost, I think the most accurate description is "desperate". The cost of two years extra support ($200 first year and $500 the second year according to Dell) is more than a new PC with a new OS...
Not the best way to spend money, but I guess some have no choice...

Considering that Windows 8 will be supported until 2023, I think it is. If you think it is too much, then there is Linux.

The "trivial" software is moved to Windows 7. The "difficult" software is tested on Win8 32 bit where most of it works. The "impossible" software is kept on XP with the measures I mentioned.

Unless they have a unique arrangement, the software is licensed for use, not rented. Also, if the covernment want it, they may have access to the source code. There have been (and I assume there are) programs for giving large customers access to the source code under NDA. Whether this would give them rights to modify the code is an issue, and even more of an issue is if the government are able to set up an entity which are competent to actually fix the code. Do you think they are? Because if they are not, then why have the code in the first place?

There must be a reason why so much XP based software apparently does not work on Win 7/8?

In my view, your points do not counter the simple truth: Microsoft do not have a monopoly. We may not like that they stop the support, but the legal avenue of claiming they have a monopoly is a dead end.

Regarding your specific issues:
The majority of XP systems can easily be upgraded to later versions. In the company I work for, only approx 1% of the systems cannot use Win7 64 bit (still a significant number). For these, we use various workarounds and fixes:
- Connect the XP machines on closed networks, behind restricted firewalls, or not on a network at all. Much of the malware out there will need an internet connection to do any harm, so network restrictions will both prevent infection and reduce the impact of any infection.
- Then there are solutions like bit9 which (when properly configured) prevent unauthorized code from running on the system.
- Our company has had good success with Windows 8 32 bit. From what I have been told, most of the software which does not run on Windows 7 64 bit (our standard platform) will run on Windows 8 32 bit.

In my organization, they have found that most of the XP software will work on Windows 8 32 bit. Thet will not help if there is dedicated hardware, but for a lot of cases, it is simply about testing the software on a new OS.

Not all reasons to upgrade is obvious to the user. Like security. A lot has happened to the underlying OS since XP. The only way to back port that to XP is to ....upgrade XP to use the Vista/7/8 kernel. Which would introduce the same compatibility problems the users may want to avoid.

Windows embedded is still supported.

There is no monopoly. There are alternative OSes you can install on the exact same hardware.

Because the point is that they want to stop supporting it. Making users feel the pain for getting upgrades is one way to do it. If the patches are free, the problem continues.
And no, I simply do not believe that people would pay $100 per year for patches to XP. They may say that, but when the bill comes, they will not.

If she is willing to pay a nominal fee, why not spend that money to upgrade the OS? After all, it seems that you are willing to help move her to Linux (which is not a bad idea), so I guess you could also help her upgrade to Win7 or Win8.1?

The case is based on false assumptions.
Microsoft still provide support for Windows XP to those who are willig to pay for it: http://arstechnica.com/informa...
Case closed.

Because all information collection in space is expensive.
Additional detectors, data storage and transmission capacity has a cost. Not only in monetary terms, but more weight on the satelite add cost and reduce lifespan. Satelite to ground bandwidth is limited, so collecting useless information takes capacity from useful information.
Why use valuable resources on collecting information that is already collected and available from commercial sources?