N100 is faster and more efficienct than any Ivy bridge E3. At idle the Xeon draws roughly 20W more, which works out to $30USD/year at the national average electricity prices. That gap widens as the load increases.
I can totally see why someone who doesnt need expandability would choose the cheap mini PC.
When I first got into homelabbing as a hobby, I built a massively overpowered server because I was highly ambitious.it mostly just drew power for projects that didn’t require all the horsepower.
A decade later, I like NUCs and Pis and the like because they’re tiny,low-power, and easy to hide. Then again, I don’t have nearly as much time and drive for offhand projects as I get older, so who knows what a younger me would have decided with the contemporary landscape of hardware available to us today.
A decently powerful Server is nice, when you need it. Having some modern APU for decent en- and decoding performance is great.
There are tasks that benefit from speed, but the most important thing is good idle performance. I don't want the noise, heat or electricity costs.
I'm reluctant to put a dedicated GPU into mine, because it would almost double the idle power consumption for something I would rarely use.
Even my old GTX 970 can throttle down to like 10W while still being able to display and iirc also h.264 decode 1080p60, let alone putting it in a mode that at all matches S3/suspend-to-ram via PCIe sleep states. I'm pretty sure laptop with extra dGPUs normalized aggressive sleep of the power gating kind for their GPUs to keep their impact on battery life negligible (beyond their weight otherwise being used for more battery) until you turn on an application that you set to run on the dGPU.