I am a ways off from a large tank (need a new place to live first), but it has been on my mind. Since i know what i'd like, but not what i'll end up with, i haven't settled on any real specs and theres always a few new ways of looking at things. I might also be overdoing things.
I have tended to dislike the idea of special purpose filtration, with a couple of exceptions. For instance, saying this filter is just for mech filtration while this one is just for bio is kind of hard for me, fish poo doesn't recognize that I am sure. I would much rather integrate everything into a linier progression when possible.
the current setup in my mind is two sumps of the same design. I'd rather have two smaller sumps to maximize the benafits and minimize the chance of getting wiped out if just one component fails; I feel rather then one large sump. Knowing what I like in fish and my penchant for overkill, I would be looking at 33% or more of tank volume for sumps. first chamber would be progressively finer mesh filtration leading to a fluidized Kaldnes media chamber in the secound to a refugum in the third with live plants. depending on final specs, maybe even a 4th mixed chamber of one or two other media types, probably primarly submerged types before exiting. each sump would have its own heaters and uv sterilizer. The biggest concern of course is space so they would virtually have to be specially made to make maximum use of avaialble space.
chamber one of course is mech filtration with progressivly finer mesh sizes. Easy to clean and monitor and realitivly cheap.
Originally I had wet/dry bioballs planned for chamber 2, but after reading about the kaldnes media, I think i'd prefer to try it out. it might be expensive hype, but a lot of people swear by it for heavy dudty use and the logic behind fluidized media seems sound. the big fear here of course is if the power goes out the bacteria has a shorter lifespan then a fully submerged type. It does make me want to be sure I have some capacity in longer lasting wet media in a pinch.
chamber 3 derives partially from the fact that I am just a big beleiver in plants in a healthy system. Anachris, hornwort and pennywort are all heavy duty nutriant sinks and anachris has the added benifit of breaking down calcium carbonate (which is heavily present in my region) when co2 deprived. part of the reason for doing two sumps is here in that with two sumps i can run the lights on alternating clocks so there is always one sump doing photosynthisis.
chamber 4 is still a thought, but it might be worthwhile to do an extra layer of filter material here simply for submerged media.
Now, for my exception on special purpose. I would certainly consider setting up a high flow canister like an fx5 in addition to the sumps. the biggest reason here would be if I decided to go with peat in my filtration system (I am a big fan of peat in the filtration) or needed carbon in it. I especialy don't feel the sump design is condusive to the best use of peat so a canister filter makes a good option here with its forced confined flow. of course, extra media in general is not a bad thing I think.