Suggest you also check (if only for your own info ;-) ) the legalities. I only ever once wanted to do this and that was in Nambia where it is a BIG no-no. There's even a section in the local air law dealing with this. Guess too many people are tempted / doing it anyway there, as Avgas is an issue, to put it mildly.
This post is a response to both this thread, and the other thread about 91UL...
Yes, I have practical experience carrying fuel in jerry cans. I avoid if at all possible. Only the very best quality cans don't leak a little, and the pressure changes just aggravate this. I have ferried with jerry cans, where there was no alternative - long legs in an MD500 helicopter and a few floatplanes, where it was necessary to land mid leg and refill.
The airfield where I used to keep the 150, had no fuel, so jerry canning was the only way to get fuel into the plane, unless I flew it away for Avgas. I got used to hefting and pouring 25l jerry cans into a high wing aircraft, but it was never a good idea - I spilled a lot over the years. Then, and still now, it is rare to find an airfield which sells Mogas around here. I don't know what others do, but I have a proper 1300l Mogas tank here. It is diked, grounded, has a pump and filter, and remote emergency shut off.
It is wise of regulators to limit the amount of fuel being transported in jerry cans. In a collision they are a hazard in many ways. I see people filling them in the back of their pickup trucks at the filling station - against the law, and a real hazard, as they are not grounded.
There was talk of ethanol, I'll reply that on the 91UL thread....
As I always fly with pee bottles I am in agreement re keeping a little bit of air on there.
The relative advantage that pee bottles have, of course, is that they are typically filled when you're flying at altitude. So they will need to deal with an increase in ambient air pressure eventually, not a decrease. Which is a lot easier.
On the other hand filling a pee bottle to 100% capacity may be a messy affair and for that reason alone I'd stop at about 75% (if able). ;-)
As I always fly with pee bottles 
I am in agreement re keeping a little bit of air on there.
Avgas expands at 0.1% per degC and one could easily see a 20C change i.e a 2% volume expansion. With negligible compressibility, the force available to burst a solid container (with no gas buffer) will be massive.
As regards how much gas space should be left in, consider the worst case temperature change. Say the post-fill volume increase of the liquid is 2%, and if there is 4% gas in there (i.e. you would see a rise in the liquid volume from 96% to 98%) the pressure in the container will double (PV gas law). Doubling the sea level pressure amounts to 15psi net pressure inside the container, which is quite a lot. My guess is that plastic cans expand a lot anyway so are OK even if 100% filled, and anyway any rectangular container obviously has a lot of volume elasticity.
Presumably, one would not use a pitot tube with a "bladder" type of ferry tank, because it would just keep it fully inflated the whole time?
Only if the dynamic (pitot minus static) pressure would be more than the weight of the top of the bladder itself. I have no idea how to calculate that. But I agree - why would you?
In fact, it's probably marginally safer not to vent a bladder type container to the atmosphere. A container filled with 100% fuel cannot ignite (due to a spark from static electricity) as there is no oxygen available.
And looking through the Turtle Pac website, I don't see any pitot or static connections anyway.
I guess this means that one should only carry 100% filled jerry cans as liquids are not compressible and thereby don't expand when ambient pressure decreases.
True, but on the other hand if the temperature rises, liquids will expand and if your container is filled to 100% with a liquid, there is no space to expand to. So the container will burst or at least start to leak through the seals.
So I'm not sure whether filling to 100% is a good idea.
My gut feeling tells me that you should have your containers filled to about 90% (so there is room for thermal expansion of the fuel, but there is not a lot of air to expand in case of pressure changes due to altitude) or 0% (means completely empty, and then you can just leave the caps off). Worst case scenario would be a container filled to just 50%. But that's gut feeling only, not backed up by real life experience or calculations.
And if I may make a stab at the reasoning why 90% is better than anything lower: If your container is filled with 90% liquid, there's only 10% gas that can potentially expand. Suppose that the ambient air pressure drops by 10%. To fully equalize this, the container only needs to "bulge out" by 1%. (10% x 10% = 1%). Whereas if your container is 100% gas, it needs to bulge out by the full 10% to fully equalize. 1% bulging out is something that I think most containers can handle, looking at the structure and stuff like ridges and small bulges in the side walls, but 10% not. (And I agree that full equalization is not needed, as the jerrycan is supposed to withstand some pressure differential. But the principle still holds.)
Anyone with more practical experience and/or theoretical calculations?
Jason has it right - Turtle Pack. These are an excellent product, and tick all the boxes for safety (for as safe as one can be carrying fuel in the cabin!). I have tested and approved them for several ferry installations. That said, if you are thinking to pour the contents into the wing, rather than pumping into the system, the Turtle Pack could be a bit awkward.
In a pressurised aircraft you can have a feed into the tanks without a pump!
Otherwise, if you go from FL0 to FL90 or above, your containers may burst.
Good point, I could theoretically climb all the way up to FL200 and back down to sea level during such a trip. I guess this means that one should only carry 100% filled jerry cans as liquids are not compressible and thereby don't expand when ambient pressure decreases.
You will need an "approval" for the connection into the fuel system, which obviously has to be permanent - at least to the extent of a T-piece with a valve, or however it is done.
Turtle pac had a booth at the AERO in Friedrichshafen this year and showed such modified fuel caps. From the conversations one could have come to the impression that "not all" pilots seek the proper approvals 
I am sure you don't need any certification paperwork for a container carried in the aircraft provided no structural limits etc are exceeded, but was wondering whether the "typical pump man" will be willing to fill up a ferry tank, without you waving a piece of paper in front of him.
You will need an "approval" for the connection into the fuel system, which obviously has to be permanent - at least to the extent of a T-piece with a valve, or however it is done. AIUI, on the FAA scene, an A&P can do that. If I ever did a flight with a ferry tank, I would try to leave the connection in place afterwards, in case it comes in handy one day - though obviously would not want something in the cockpit that gets in the way.
ensures that the container is vented to the atmosphere without the danger of the fuel syphoning out
Presumably, one would not use a pitot tube with a "bladder" type of ferry tank, because it would just keep it fully inflated the whole time?
Never done it myself, so can't speak from experience, but I can see a few issues that need to be solved.
Filling. Make sure you do this outside the aircraft. Spills inside the aircraft are bad news, if only for the smell.
Crashworthiness. You're talking about quite a lot of dead weight. Make sure they are well stowed and secured. (And obviously check for W&B issues.)
Air pressure. The pitot tube that Peter was talking about doesn't so much pressurize the fuel so that it flows into the engine without the benefit of a fuel pump, but ensures that the container is vented to the atmosphere without the danger of the fuel syphoning out. Otherwise, if you go from FL0 to FL90 or above, your containers may burst. With a turtle pack or other type of flexible container these issues are not present, obviously.
(And actually I'm not sure which rigid containers are most susceptible to bursting: those that are empty, those that are 90% full or those that are 100% full.)
Refueling from these containers - static electricity and sloshing. Make sure this is done safely with no spills and such. The easiest method I've seen is by making sure that the jerrycans are above the fuel tank level (set them on top of the wing for instance), and then use a hose with a one-way valve at the jerrycan end. You jiggle the hose up and down in the jerrycan so that it becomes "primed", and then the syphoning action will do the rest. Because you are not agitating the fuel, and only expose the fuel to the outside air inside the jerry can or inside the fuel tank, the chances of static electricity creating a spark and igniting the whole shebang are minimal. And since the jerrycan sits on top of the wing throughout the process, there is a natural bond in place already.
Can you get fillups if you have a ferry tank, without showing a ferry permit for that flight?
I would think that a ferry permit would only be needed if the ferry tank is a "permanent mod", for instance plumbed into the fuel system of the aircraft, but for which no STC or anything is available, OR if the weight of the tank and contents are so that you would depart overweight. The ferry permit then allows you to fly outside the normal certification limits, for some specific flights, subject to certain limitations. But if your ferry tank is STCd and you are within the W&B envelope, I don't think a specific ferry permit would be needed.
