This answer redirects to OpenXtra that suggests server rooms shouldn't run too cold. Google also suggests not going lower than 55°F, but doesn't go into details regarding why.
Given that:
- CPUs do work completely fine while in sub-zero temperatures.
- Hard drives will fail if going into uncommonly low temperatures (which won't be reached if living anywhere between northern Norway and southern Argentina).
- In general, electronics won't be ruined by cold, but by condensation and humidity.
Why would it be bad to run my server into an unheated basement as long as the temperatures aren't extremely low (-10°C to +15°C) and the air is reasonably dry?
All systems have an operating range. It really depends on the vendor/hardware. A standard HP ProLiant has an operating range of 5°C to 40°C (41°F to 104°F). HP also has ruggedized ProLiant variants with an operating range of -10C to +50C (14°F to 122°F) (Assuming SSD storage).
First off, its NOT bad from the servers POV to run rooms colder, within reason.
The links you have imply, or downright state the reason - cost. Keeping servers cooler costs more money (and wastes more power with its related environmental impact), and does not significantly increase the reliability of the systems during their useful service life.
Remember that computers are typically designed to work in places people do - servers, while differently focussed, are very similar in their design. Electricity is a major DC expense, and cooling is a fair chunk of that. The problem is amplified as well - cooler room means more aircon hardware using more electricity, bigger ups's and generators, more onsite fuel storage. More maintenance and less space for cabinets.
There are other, lesser reasons to consider as well. One that sprongs to mind is humidity. The colder the air, the drier it is (that's why despite all the ice, the Arctic and Antarctic are deserts). Less humidity means greater opportunity for static discharge, which can damage hardware.
Another could be stress during - the delta between typical temperature and temperature in case of failure is larger the colder the environment is. Higher differences in extremes is more stress is greater likelihood of failure.