Finally got my T26xx empties to play with and pulled one apart.
While I'm almost certainly going to stating the obvious and/or repeating some previous intel', I'm just going to go with it rather than trying to refer back and forward 15 times, so apologies in advance.
Think of repeats as a useful summary 
*note: I'm going to refer to the Dribble refilling method as DM throughout for speed
Ok... first thing to note is that these cartridges have no window like the PGI-9 so there's no handy means of referencing your ink levels visually. There is the little prism in the base and I suppose it might be possible to use that if you could figure out the differences shown between full and empty. Use of jewellery scales obviously helps you resolve the issue of how full the cartridge is with ink but what it won't allow you to do is figure out how much air might be in there.
At this stage, I'm not too sure how important the air issue could be for DIY refillers using the DM but only testing is going to help with that.
Disassembly
Next thing, again, rather obvious, is that the cartridge side is sealed on good and tight... It's an absolute pig to get off and in the process I punctured the bag. Based on what I've learned so far, disassembly is not going to be required for any DM refilling but worth mentioning it for anyone thinking to give it a go. Expect the cartridge to be knackered afterwards.
This image shows the cartridge taken apart and two items in particular I was interested in relating to venting:
1. Case vent hole.. This is the only way air can enter the cartridge unless the cartridge has suffered damage to the seal
2. Bag vent mechanism..
Design Features
There is a single vent from the external to the internal cartridge body and air flow into this is controlled by the vent matrix in the detached side panel.
To show what happens I've traced the matrix path and roughly superimposed it over the internal cartridge.
Basically, the plastic film
(shown in earlier photos from Mikling) is sealed over the matrix in the cartridge side panel. The seal extends across from the vent (1) until it reaches 4. Here the film is left open and it's basically the other end to allow air in and out (3). As with all other cartridges it compensates for ink usage although in this instance it's the bag deflating.
The bag vent (2) is, as suspected, triggered by the bag deflating until the metal plate triggers the plastic arm that sticks out under the plate. When triggered it allows air into the bag. However there's no secondary inlet for the air to get into the bag, and (2) only has access to the air already within the cartridge body.
Whether by design or not, the fact that the bag vent does indeed trigger when the bag is empty, does mean that any commercial operations looking to flush, vacate and refill, are going to have a tough time making it work.
Actually thinking it through further, no it's easily resolved (I think!)...
i) Seal the vent hole
(and shallow channel leading to cartridge side) on the side panel (aluminium tape?)
ii) Pull a vacuum on the cartridge, via the outlet
This will draw air/ink out of the bag, trigger the bag vent and draw any remaining air from the cartridge body into the bag and out through the cartridge outlet.
iii) Partially refill bag with cartridge in an upright position
(so gravity isn't readily providing ink to be drawn out through the bag vent to mess up the internal cartridge outside the bag).
I suspect the amount to fill at this stage would need to be figured out carefully as the vent will be open and drawing anything it can out to fill the vacuum outside the bag. So some testing needed here.
iv) Unseal the vent hole (the trigger should no longer be active).
v) Complete refilling
Using the same process to flush out inks is still going to be a pain but at least it's feasible... and doubtless someone else can finesse that somewhat but as a first draft, not too shoddy
Note: The process above is intended to deal with the problem commercial refillers are going to face (flush, purge, refill) but if air ingress to the bag turns out to be problem for DIY refillers using the DM approach then a flush clip
(as we've done with PGI-9, etc) should solve the problem and a simple enough syringe based solution is already simmering in the old noggin
As for future testing, I'll need to source some more of these cartridges and try to get one apart without puncturing the darned thing and then stick a window panel over it instead so I can see what happens. In particular I'm interested to see:
a) Whether normal usage "empty" does indeed mean the the bag is filled with air
b) If true, what that means for dribble refilling
c) If it creates a problem, how to mitigate it
(Likely using the method given above).
Been an interesting exercise so far so hopefully the above contributes to the discussion to date...
