50-70% is the best loading for efficiency
That isn't true. Power supplies "peak" efficiency range from 30% relative load to 70%, depending on the model.
And in PSU of modern topologies, the efficiency deviation between ~20% and ~80% load is about 2% max. This is mostly negligible. Even more negligible when you consider idle loads, where the higher wattage power supplies are generally less efficient.
Different model power supplies will have different efficiencies. (made up example) An EVGA G3 may be more efficient whilst at 90% load compared to an EVGA B3 when it's at its "peak", simply because the EVGA G3 is an overall more efficient power supply.
If efficiency is really a concern, choose an overall efficient power supply.
Cybenetics Efficiency ratings are a great resource (like the 80+ rating, more much better and newer methodology). Wasting time/money messing around with wattage load will get you no where with efficiency.
Relative load has nothing to do with thermal regulation. What matters is the real load, eg. 300w. Two power supplies in the same series (and platform and fan), for example the EVGA G3 550 and EVGA G3 750, will have similar efficiency and layout. 300W on both power supplies will produce the same amount of heat, even though the G3 550 is at 55% relative load while the G3 750 is at 40% relative load.
Makes for a quieter psu if nothing more as the fan won't be at max speeds constantly.
In most cases it's the opposite. In most power supplies, including alot of high end models, the fan speed is controlled by temperature. Since we know two power supplies of the same series and platform will produce the same temperatures, we can infer the same amount of heat needs to be dissipated.
However, as the higher wattage power supplies need to be capable of dissipating potentially more heat, a more aggressive fan profile is used.
This is evident in Cybenetics data, comparing the
G3 550 with the
G3 850. During idle load (60w DC), the 550w version's fan spins at ~1200RPM while the 850w version's fan spins at ~1540RPM. This is also evident at other loads.
Oc 8700k and 1080ti uses closer to 400w generally, give it take depending on use. A 1070ti/1080 uses closer to 300w+. A 8700k can easily exceed 200w by itself under OC. Add a couple of drives, ram, RGB, fans etc and the pc itself can see 300w before any gpu considerations.
https://www.techspot.com/review/1515-msi-geforce-gtx-1070-ti/
They use an i7 8700k Overclocked + 1080ti, they recorded a maximum of 391W from the wall during intensive gameplay. Due to PSU efficiency (about 90% with their power supply) their system really consumed a max of 350w.
If you believe the Corsair TX650M will be quieter and more efficient than the Seasonic Focus 550w purely because of higher wattage capability, that is wrong.
The
Focus 550 measured 10.5 dBA at 330w, while the
TX650M was 39.8dBA at 325W. At idle there was an obvious difference. At the same ~330w load, the Focus was only 1.2% less efficient, and at idle loads the Focus is 2.5% more efficient, which is basically indistinguishable to most peoples' power bills.
So OP's system will only be using ~63% of the PSU capability during gaming load, not a constant 100% as the manufacturer guarantees.
Add a couple of drives, ram, RGB, fans etc
SSDs consume under a watt, HDDs consume 5 - 10W, RGB and fans only a couple watts max.
There's nothing to worry about, wattage wise, with the Focus 550.