The statements about shooting f/4 have to be taken with a pinch of salt if you go on a dof calculator you find that
The Rx100 mark II of the op when at full zoom with f/4 has a depth a field of 0.4mm at 10cm which is nothing at f/11 has 1.1mm at same distance
A micro four third on a 60mm lens at 19cm which achieves full macro has double the dof of the rx100
An SLR full frame with a 105m lens at 30cm which is when you get 1:1 has 3.6mm
I have shot the Rx100 a lot with pygmy also in video and f/11 is the way to go for super macro. Generally always between f/8 and f/11 for macro while f/5.6 is for wide angle and portrait work
Wider apertures I have only used in video due to the fact I set ISO max to 800 and the corners are not sharp until f/5.6 on rare occasion for stills I have used f/4 and wider only for bokeh
In terms of strobes the inon z240 or the sea and sea DS or DS2 work fine obviously if you get some Z330 they will go even further with you. S2000 or similar are not a good idea in my view
The point of what I posted is that for the same DOF you can have a wider aperture with a 1" sensor compared to a bigger sensor, So for wide angle which is most demanding on flash power as subjects may be further away generally you can shoot f4-5.6 range and get the same DOF as you would with f11-16 on a full frame. Because flash runs purely on aperture it means you don't need as powerful a flash. I wasn't suggesting for a second shooting super macro at f4, shooting at f8 gives about the same DOF as f22 on full frame, which also requires less flash power than that needed on full frame. Because the strobes are so much closer power is less important, even though you are stopping down. So if you are staying with a 1" sensor camera something like two S-2000 strobes can be enough and also easier to get in close to your port if needed and overall more balanced with the small housing. I note the OP mentions a Sony housing, assume it's acrylic and probably positively buoyant, having two heavy flash above the buoyant housing may prove a little unbalanced and will want to float housing up. Having said that the INON strobes are all quite close to neutral buoyancy, other options I'm not sure about.
Stepping up to a m43 setup or larger sensor, you may need to stop down some more and the housings are getting to the size that two Z330 makes more sense. If you think you might upgrade to a larger format in the future then the bigger strobes may be worth it. If you're on a budget you can get two S-2000 for the price of one Z330 more or less which may be a good option for some.
As far as arms go, weigh your housing underwater with a luggage scale assuming it sinks. INON publishes underwater weights of all their equipment, you can quickly work out the UW weight by adding the weights together (subtracting items that are positively buoyant). and work out how much buoyancy or weight to add to the rig. Link to the table is here:
http://www.inon.jp/products/armsystem/weight_table.html