Wow, that sounds very over-complicated. What are you trying to achieve?
My fuel pump is the Huco type, the most it will ever take is 5 amps or so, typically less, in fact hardly anything on average as it works in pulses. The wiring is quite tolerant of this type of load as it isn't a continuous load which applies heating effects to the wire, think of what over time this would be as RMS or "average" power and it adds up to very little. In fact wiring will have several ratings, continuous and different time periods. Looking at the AWG standards, most wiring will run at several times rated capacity for 10 seconds, so your pump pulsing every few seconds at a very modest 5 amps isn't going to make any difference. Not that you should get anywhere near these limits, just by way of illustrating the problem you are dealing with; wire has resistance which causes heating when you put current through it, your pump really isn't doing that much of the time.
No relays, nothing to do with starters or tachometers, just on the ignition switch with it's own fuse taken off before the fuse box. Seems to have been OK for the last 4 years. The only thing I added was an inertia safety cut-out switch which is just wired in series. I shouldn't need to spell out why this is a really, really good idea... Disconnect your pump from the carbs and see how much fuel it can shift.....
Different story with the fan of course as this can run for extended periods, however you can mitigate this. Because my installation has two speeds via a resistor pack, you don't get quite so much of the massive transient currents on start-up - the resistor limits this. Even when it changes from low to top speed, because the fan is already spinning, you have "back EMF", so the draw is reduced. This is merely wired to the permanent 12V (not the ignition) with it's own 20A fuse so it can run-on after I stop if necessary. I ran it for a while with a 15 amp fuse and this was OK for most of the time but would eventually blow. If you read my thread, I reckon with the resistor in place for the low speed (and yes, you really DO want it to run at low speed most of the time) the load drops to about 8 amps continuous which is really only half the power of the rear-screen heater - full power is around the same as the rear screen heater and both are incredibly unlikely to run at the same time. (in fact my screen heater doesn't work anyway). This is why I'm not really concerned with adding relays and taking a new feed. Of course it will depend on the exact hardware you are installing.
I had a manual override but I removed it as it was completely unnecessary in practice - well, no 1980s or 1990s vehicle with the same type of thermo-switch had them so why should it be necessary? It's completely reliable in operation as proven in 100s of millions of vehicles.
I recently had a re-cored radiator and I have an expansion tank. My conclusion from this is the P6 cooling system is actually rather good if kept in good condition which you'd expect from a large radiator. The omission of the expansion-tank as standard was a very bad cost-saving idea. In fact if you are planning to have a thermoswitch I would say it is a "must" to keep the switch in contact with the water in the most optimum position (where the water enters the rad). At 28 degrees it is really quite difficult with the car idling to get the fan to kick-in at all (95 degrees). In fact I'd say compared to a modern car, the cooling has far more capacity. What my new VW has is better control of water flow and the fan which allows it to use a smaller and lighter radiator. Remember, modern cars run leaner and hotter than old cars (even ones with V8 engines) because that's how they control emissions and boost efficiency.
It is of course extremely wise with old car electrics to be cautious. In a modern context they are very underpowered. Ideally I probably would have run the fan from a new feed from the alternator but I still wouldn't use a relay, only an independent fuse - as it this is just a case of running another wire and I should get around to it. I think the trap most people have fallen into is that because it's a big engine they've added a megawatt fan. This ISN'T necessary. You have a radiator with a very large surface area, you just need to make it work as efficiently as possible. As my playing about proved, it takes less than 100W using the Spal fan to cool a V8 P6 in any circumstance you are likely to encounter, it just needs the fan to sweep as much of the radiator as possible. Those who bought fans which can draw 40 amps or so have just made more problems for themselves IMHO.
Think about basic cooling principles, the rate of cooling is proportional to the excess temperature. Therefore what you want is to cool the hottest parts of the radiator where there's the biggest temperature difference. If you've already cooled a part of the radiator to half-way between the engine and the air then you are cooling at half the rate so throwing more hurricane-force air through isn't going to make so much difference as making sure the really hot parts have at least some air moving over them. Then of course you generate a pressure as the air resists moving through the matrix which the fan has to work against further reducing efficiency - these effects aren't linear either, they increase with the square and cube of airspeed. Basically lots of effects kick-in which work against efficient cooling by brute force. What you want is big area and modest airflow - which is of course how the standard engine driven fan is setup in the first place. For sound engineering reasons it would seem