For 12ga wire vs. 14ga wire I doubt the difference in cost is as great as all that. But I could be wrong; can you show us your math? Here's mine: With a cursory search, I find TXL 14ga wire for 18¢/ft and TXL 12ga wire for 24¢/ft here. If you use 50 feet of 12ga or 14ga wire in the whole project, which seems like a fairly large overestimate, at those prices you would spend $9 for 14ga or $12 for 12ga. I have a hard time regarding that $3 difference as a "significantly higher expense". Or suppose you want to go unnecessarily far toward the high end of wire specification and use silver-teflon wire; I find 14ga for 95.8¢/ft and 12ga for 80¢/ft (no, those aren't reversed). Where's the "significantly higher expense" of using 12ga here?
I'll have to look where I got the wire--but not there. Part of the problem is where I did order, you had to buy in quantities of either 100 or 250ft minimum, which is obviously far more than necessary--but with a significant price difference, that extra quantity adds up. I think it was nearing twice the price of 14GA, but the purchase was over a year ago.
I also had problems with connectors. Some of the stuff I've used for light plugs and fuse systems (weatherpak type) won't take 12GA at all, and max out at 14. That's another issue that added to it, which is admittedly more of a problem with component choice than the actual wire.
As an example, what I've seen of the weatherproof Hella 90MM modules, the connections would not comfortably (and may not at all) take 12GA wire. 14 might even be a stretch in some cases.
I would be interested in your test results from the different relays, I used Hella 40a relays in my build and used one relay for low and one for high, I used 12ga wire throughout except for trigger wires, I have almost zero loss in my harness (battery is in the trunk and I pulled a 6ga positive and ground under the hood un Power all my extras...)
What were you findings with different relays? I have plenty extras, I could send you one if you didn't already have one to test..
I was testing mostly full-sized Tyco relays against some micro ISO relays. I have an Excel spreadsheet that details all of my findings, but I don't have it here. I'll have to see if I can find it. I tested one and two bulbs per relay. As expected, the voltage drop was higher with two bulbs per relay, but significantly moreso on the micro relays. So much so that you could get away with a single average car headlight bulb on a micro relay, but adding a second was (IMHO) too much voltage loss.
The factory wiring had surprisingly small amounts of voltage drop--but the circuit was fairly well-designed, and the wiring was very new. The circuit didn't look much different than what we'd build in our own relay harnesses. I'd guess the wire was 20GA, but it had short runs to ground (using the chassis). Surprisingly, the least drop was the furthest from the batter/relay.