If you don't run slowly at high engine temperature in really high ambient temps with intense sunlight on the tank, the pump will probably not fail. A lot of the failure reports come from Australia and the American southwest, though not all.
I have to preface this with the statement that this is only a theory of mine based upon my pump, a couple of other pumps I've gotten my gritty mitts on, my limited testing, and that I may be full of shit. There, that ought to keep the lawyers at bay.
I did some testing after my pump failed. What I found was that the pump membrane appears to be made of a material that behaves as if it were thermo-setting plastic. At 150°F the membrane, in the presence of 10 inHg vacuum, it will form to the domed shape of the pump cavity.
It will recover if it is taken out of the pump and heated to a bit higher temperature. Obviously, since it is unlikely you'd be able to heat the membrane in the pump without there being vacuum present, it won't recover in the pump. Once failed, it's kaput.
I was skeptical that the temperature inside the tank could rise that high, but proved that it could if the bike is sitting in ambient temperatures of 100°F with direct sunlight and the engine is at operating temperature. Remember that the tank is in very close proximity to the head and almost completely envelopes it. Running slowly makes it worse, as there is no air flow under the tank except for the hot air coming off the radiator.
I did tests with both black and translucent vessels, thinking that the translucent tank would not be susceptible to the 'black body' heating. It was only a few degrees cooler, and I've heard reports of pump failure in translucent tanks.
The problem with the failure reports is that people chalk the failure up to contamination and are really pleased that IMS will send them a new pump for free (by the way, these pumps are used in ultralights, personal water craft, and snow machines and they never
fail except at extreme age with contamination). They never do a post mortem. I got a used pump that was being held as a spare. When I disassembled it the membrane was domed.
The insidious thing about these pump failures is that, unless you regularly run your tank down more than 8 gallons, the pump can fail and you'll never know it. That's is what happened to me, at 70 mph in the fast lane of the 405 freeway in Los Angeles. It just ran out of gas, but there were 7 gallons left in the tank when I went to fill up. I think few people regularly run the things down much more than 5 or 6 gallons.
I've run a lot in Nevada, Utah, and some Wyoming/Montana. Having a range of 450 miles is nice
I don't have any pictures of the finished job, but here are some in-process photos: