Sorry for the delay, I'm on night shift at the moment and pretty busy but Sunday night with everything steady has given me an opportunity to sit down at the computer.
Regarding preferential flow, this is going to be present any time you have a wall and a packed bed. A 2/3 diameter dome can will reduce amount of preferential flow, as this flow down the sides now has to come across horizontally into the area under the dome. This comes from the school of thought - which is actually a very correct, well-educated school, i.e. Weihenstephan - where slow, gently, drawn-out flow through a bed results in the highest efficiency. It should be applied during mash out conditions, but should be completely ignored during step changes where high flow is your friend. You can hopefully appreciate where a 2/3 dome is a good choice (increased lautering efficiency at mash out conditions) but is a big facepalm/d'oh when you are trying to get high flowrates when stepping your mash.
So, with that all considered, assuming a flat bottom mash tun...
- A braided hose offers the worst of both worlds: cripplingly low flow and big sacrifice in lautering extraction (efficiency).
- A full diameter false bottom hard against a flat bottom is probably the next step up, but the zone of compaction will still be pretty focussed right over the draw-off nozzle, which will limit your flow but might slightly increase your lautering extraction.
- A 2/3 diameter dome is the next best, with a more even compaction zone but still has a compaction zone over a reduced overall area
- A very close to full diameter false bottom with a voile bag is next best, as it permits the highest possible flowrate with a very small (difficult to measure...) sacrifice in lautering extraction due to the preferential flow down the sides
- A sealed, full diameter false bottom with high surface area (lots of holes, not much blank space) and a good few inches of disengagement space off the bottom is pretty well as good as it gets, but is going to be hard/impossible to fit into a keg unless you have a split design and silicone tube around the edge when you assemble
- The motherflippin ****-hot best solution is a full diameter, sealed false bottom with a 1:1 diameter to height or greater, a few inches off the bottom, with a circular baffle plate of 1/3 diameter an inch off the bottom to further break down the velocity and give maximum bling, efficiency, flow and bragging rights
A bit of SS duct/pipe with cone welded onto the bottom matched to a full diameter false bottom and voile bag for backup would be awesome. A full diameter dome on top of this cone would be pretty close to the full diameter flat with a stand, but a 2/3 diameter with a cone bottom would probably be worse than a 2/3 diameter on a flat bottom as your preferential flow down the cone sides is still there.
With the keg, you might struggle to get a full diameter false bottom in unless it's split, and a 2/3 diameter false bottom would actually be pretty bad given the very high bed loading (kg mash vs false bottom diameter) so your flowrate would actually be pretty low. A 4800W, 5, 6 7kW element would be great, but it would be off most of the time due to the low flow - i.e. the wort coming in over the element is pretty hot already, so the controller keeps it off/low most of the time so this would be a waste of money. Having said that if you make a good mash tun choice (duct/pipe with cone, full diameter false bottom) a 4800W would be your best bet, with a 3600W being ok but might limit you a bit. Run at max flowrates during ramps, and back this off through each step and during mash out/sparge to get the best efficiency.
Baffle plate is optional but that further spreads the velocity out to ensure you get a confident distribution over the full width of your bed without slightly higher velocity and compaction coming up through the middle zone.