This is a good point. Matt, are you doing any physical inspections, cleaning out dust, etc? You might want to add a CPU test of some sort to your toolkit, ramp up processor usage intentionally, and watch the clock speed to see if the processor is slowing down due to thermal protections. Some sort of disk test would also be good, since a slow hard drive is going to slow a computer down too. Some computers also just don't age well at all. Even though a computer comes prebuilt doesn't mean it was designed well. I can't tell you how many dead Dell XPSes my two previous companies went through since it was so many. A combination of poor office cooling during off hours, high workloads, and a bad placement of the SATA controller directly behind the heat exhaust from the CPU caused countless failures, leading to both companies abandoning the XPS line for different models. Very few of the systems just outright stopped one day, most degraded over time. One I remember from a programmer next to me went from 10 minute compiles (standard at the time) to 50-60 minute compiles over the course of about a month. IT reimaged the machine, and it still took the extended amount of time for his compile to complete.

As for disk imaging, I don't have any good suggestions for what you would need. The IT departments where I have worked tend to use enterprise only tools, built around having just a few standard types of machines minimizing drivers needed. They also are more set up towards recovery over the network via PXE boots.