To whom are we speaking?? Add a signature please.
Originally Posted by GypsyRain
We're talking 4,400Wh/day TOPS. Eventually (ie in a few years) we'll want to take that up to around 15,000Wh/day. When we add the fridge, washer, etc.
The former number is not a small value, the latter is fairly large (as these systems go). Doing some rough math and assuming you'd like to have enough power stored to last two days, you are looking at a massive battery bank - far larger than you will find on nearly any rv/bus. That will almost certainly require lithium (and LOTS of dollars - like $15,000 or more). If done with typical lead-acid batteries, you are looking at a battery bank of about 22 batteries which will weigh something around 1500 lbs. That is extreme and probably means that you need to do some experimenting to figure out what you really need for power (probably over-estimating).
In my opinion, the shortest distance between two points is a straight line. So, it would be ideal if you designed your system now and then added components as budget permits. However; many folks simply do not know where the system will end up (what they need) so doing this is simply not possible.
For the panels, you can certainly start small and add later. However; you need to have a plan in mind. For example; starting with 12V panels and a relatively dumb charge controller and then later adding high voltage panels would not work. You already appear to be looking at 'better' panels so will likely choose/require an MPPT charge controller. Get one with the capacity to match where you think you'll end up. There is an "ideal" array to lead-acid battery sizing rule of thumb - it is the ability to charge the battery bank at 13% (ah capacity - minimum of 5%). However; don't get too hung up on this as it is not always realistic on an rv/bus and I've yet to see real world verification of the value of this for our type of use (i.e. the batteries will last twice as long). This is not an issue with lithium.
As long as you buy components with more capacity that you need, you have room for growth. However; that is no guarantee that the component will be the correct one down the road.
A 24 volt battery bank is the better answer (than 12V) if you plan to have large capacity and consumption (and you said you do). It certainly does not hurt on smaller systems and, in fact, may be a better option (electrically) in every situation (over 12V). 48V is even better (electrically) but, as near as I can tell, the number of components (inverter, water pump, lights, etc...) that will work with 48V is relatively small (fewer options) which, IMO, makes it a somewhat less attractive option. The negative is needing 12V power as so many devices require it - and to do so requires a step-down converter. Not a big deal or expensive but an extra component. There is no magic number to know where a 12V battery bank becomes less than ideal. Your solar array and charge controller factor in here. Most MPPT charge controllers are rated for a maximum output of X amps. This is regardless of the voltage. So, the TS-MPPT-60 from Morningstar (for example) will output a maximum of 60 amps to a 12V battery bank, 60 amps to a 24V battery bank, or 60 amps to a 48V battery bank. Convert that to watts and you'll quickly see that you are doubling the actual capacity of the charge controller each time. This can result in significant cost savings, depending on the size of your array.
If you choose to go with a 24V battery bank, do it early. It affects several of the more expensive component decisions (battery charger/converter, solar charge controller, inverter, etc.).
Mixing batteries (types as well as old/new) is generally considered to be bad. However; many people do it. Ideal? No. End of the world bad? No. Thus; it would be ideal to decide what you want for a battery bank and build it all at once. But, again, 'ideal' often does not align with reality. So, building a two lead-acid battery bank today (for example) and adding four more later will work. However; you may not see optimal performance for a bank like this. Lithium is a bit different. However; unless you can state that you need very significant capacity or want to spend a big chunk of money to play with lithium, I'd say to avoid it for now. But, as already mentioned, if you go with lead-acid today and later decide you want lithium, it may affect some components (especially charge controller and charger/converter - specifically their ability to properly charge lithium, which is different than lead-acid).
All that said, there is nothing wrong using a single 12V battery for starters. It will provide for some very basic power needs and help you figure out what you need/want. In fact, used 12V UPS batteries are often for sale on Craigslist and these can be used to get you started. Of course, if you are constructing a battery box (especially one to hold 20+ batteries) and spending money on cables, it might be nice to do that just once.