The effects noted here became evident when flying small electric helicopters indoors. This is a good way of comparing batteries and their performance, if flown in a restricted space, because you can do little more than hover or perform pirrouettes. Thus the load is fairly constant and the flight time is an indication of a batteries performance during its life and allows you to compare similar and disimilar cells.
I had been using 2S packs of 1200 mAh Kokam cells to fly my 'Hornet' helicopters. When I acquired a 'Piccolo' heli, I expected to use the same packs. The latter is rather lighter and has a less powerful motor. I was surprised to get a poor flight performance and a duration of only about 1/3 of what I expected. There would be a sudden loss of power and the heli would sink to the floor. I checked several times that the battery concerned had actually been charged. When I put a given battery straight back into a 'Hornet' I got the expected performance and a duration which was what I would expect from the remaining charge.
I've noted elsewhere that my first Li-Poly (2S 700 mAh from Falcon Models) battery was bought for me by a friend. This had only been used in lightweight park flyers and would fly the 'Piccolo' well and give good duration. For a while, this was the only battery I had that would fly that model. It is still in use but eventually lost performance and would only give very short flights. Unfortunately, this was before I started logging the number of cycles so I have no idea just how many cycles it has on it, but I suspect it would be less than 40.
So, the lesson appears to be that if you get a Li-Poly battery accustomed to a certain duty cycle and then change to a different one - even if it is lighter - it will give poor results. 'Poor' in this case can mean much less.
I didn't worry too much about this situation at the time, but it is not the end of the story. I have two examples of a small helicopter (Walkera '5G6') weighing around 2 ounces and each powered by a pair of '1215' motors driven by a single Li-Poly cell.. Current draw in the hover is 2.4 amps. I have been flying these on Fullriver 250 mAh cells - thats a good 10C but they seem happy enough and give nearly six minutes flying time..
I also have a similar heli (E-Flite 'Blade mCX') weighing just half that (1 ounce) and powered by a pair of pager motors driven by a single 110 mAh cell. Current draw is 800 mA (7.5C). Flight time is 8 minutes. It seemed obvious to try this with the 250 mAh cells. This gave a weight increase of 3 grams (10%) and a poorer performance. The shock was that the heli would literally fall out of the sky at 3 minutes. This heli has a low battery sensor that reduces power when the battery volts fall below 3, accompanied by a flashing LED. When the 250 mAh cells lose the power to fly the model, the LED doesn't flash.
The obvious next step here was to try a new cell on the Blade and see if it was happy. Things were then made more urgent by the original Blade battery failing at 27 cycles. I located some 200 mAh Fullriver cells which were lighter than the 250 size and gave a weight increase of only 1.5 grams. The first flight was just over 13 minutes. Flights 2 and 3 gave similar times. This seemed fairly conclusive.
It's worth pointing out that we are talking about single cells here. To sumarise, once a Li-Poly has become accustomed to a certain duty cycle, if you change the duty cycle - even for a lighter one (something like a third in the above case) - the cell doesn't like it. Call that what you like, but I call it MEMORY.