@ molle
I found your earlier approach to the problem very convincing too. I have tried the taping and it solved the pulsing, but I did not want to have tape sticking to my airvents in an new computer. Tape glue tends to dissolve under heat and then the tape does not stick any more, which is annoying and I did not spent a good deal of money for a new laptop to fix things with tape.
Obivously it is a servo-loop problem between the temperature reading sensors and the fan control rules. The fan starts pulsing, when the temperature drops to a threshold temperature, but immediatly rises again to that threshhold, when the fan spins slower, so the fan speeds up again, cooling the temp sensor down again, which makes the bios regulate the fan speed down again and so on and so on.
By taping some airvents you reduced the airflow in some areas, so that this vicious circle is broken, but also the cooling is reduced. That certainly does the trick.
But: the controlling software should allow a proper regulation too, when the thresholds are defined in temperature areas, where this effect is not kicking in immediatly, but in larger steps (which I speculate is what the lenovo engeneers implemented in BIOS 2.52, since the temperature regulation starts at 60 rather than at 50 degrees celsius), or you would have to define a certain time lag between slowing the fan down or speeding it up or you need seperate temperature thresholds for a faster and slower fan speed. And: there are more than one temp sensor in the laptop, so you could take only those in account whose readings are not that volatile, but give a more steady temperature curve.
What i am interested in: does your fan ever stop with the old BIOS 2.50? In my experience it did not, despite the taping.