Read the thread quickly, but here are a few tips:
1. Never overclock or run anything out of spec on a system you're selling to someone and doing support/providing warranty for. They won't see the difference, and you're just increasing the odds of something going wrong.
2. I have my own IT business, and I don't think I've sold an AMD system in the last 10 years. Intel CPU and Intel chipset have less bugs and perform better.
3. Charge more! (I don't know what you charge now, but just reading your posts I know it's not enough lol) You'll lose the cheap as f... clients, but you'll make more money in the end. Or worst case, you'll make the same money in less time.
4. Like others have mentioned, dual-core is probably enough for now, but if spending an extra 100$ for a quad-core means the computer will be usable for 2 extra years, it's worth it. (example: a Sandy Bridge dual-core I wouldn't touch these days, while a quad-core from the same gen is still decent) The cheapest desktop I'd sell today is an i5-6400 with an SSD.
5. Don't be afraid to tell someone their computer is just too old, and refuse to fix it. When a client calls me about a 5+ year-old computer, and I see it's going to cost half of what a new one would cost to fix it, I tell them it's time to replace it. Believe me, I've been there many times. It always takes more time than what you quoted to fix it, and they end up with the same out of warranty old computer that'll break again in 6 months.
6. Buy prebuilts. The margins are razor-thin and the OEMs pay much less than you do for Windows, so there's no saving trying to build cheap computers. (only exception would be gaming computers) Take into consideration the time it takes to build them, test, install all drivers, the odds that there'll be some incompatibility between parts, etc. Building computers used to be an alternative, but now it's a HOBBY.
Maybe you're doing this on the side for fun, but if you're serious about making a living out of it stop being nice and be professional. (people calling at 11:30pm and you offfering to fix it for 30$, that's what a 15 year-old kid would do) Part of the job is recommending what's best for a client in the long run, and 99% of the time, spending more now saves money later. And 99% of the time, clients don't understand that, so you have to convince them or tell them to find someone else to help them.
Sorry if I seem harsh, but I was like you when I started. Always trying to find the cheapest way to fix something, instead of doing it the right way the first time. And if your client is too poor to pay to have it done right, that's not your problem. If I go to a car dealership to buy a car and their cheapest model is 20,000$, but I only have 10,000$ to spend, they won't sell me half a car. They'll tell me to wait until I can afford it or go somewhere else.