I think we already have made very valid points that Lenovo engineers should consider. And I don't think that reverse engineering the BIOS is really a piece of cake.
Let me get back with a list of questions that I'd like seeing officially answered, @Mark_Lenovo, could you please pass them to the engineering group? Some have already been asked, I'm repeating them since they may be buried many pages ago. Please, could you give feedback on these and not just throw them in the rubbish bin?
1) Could you ask to the engineering why in the world there is such a gap between 2000 RPM and 4000 RPM? It has been shown that the fan is able to run at 3000 RPM, why don't they stick that speed again into the BIOS, letting the laptop using it for the sake of a more quiet life when its not hot (that is, always). A better fan control is certainly possible and definitely needed. I cannot stop stressing how beneficial this would be! Please, ask them what they think of it.
2) Are Windows drivers or the Lenovo power manager software actively controlling the fan behaviour, or they simply reduce the power usage such that the BIOS lowers the fan speed? Said in a different way: is the BIOS the only thing that decides the fan speed, based on power drained by different components, or is Windows helping in setting the fan speed?
3) In the case that the Windows drivers or Lenovo's power manager directly set in some ways the fan speed, why don't you put that control directly into the BIOS (where it belongs), such that every OS benefits from it. Maybe you can add a BIOS option asking if a user wants to sacrifice some performance for the sake of a more quiet behaviour, mimicking the Lenovo's power manager (I think the BIOS already has some power manager options, but they don't seem to produce a noticeable effect)
4) Could you ask the engineering group to try a Linux live CD. It is as easy as to burn one and stick it into the optical drive. I do not know what software they are using on Windows to measure performance, but it will be useful to have an official word about the fan speed on Linux. I'd simply love to here: "we tried, the fan spins fast and we have no means of finding more under Linux, but hey, at least we did boot it!"
5) Can we have a better understanding on how the fan control algorithms work? I understand that it is impossible to leak the complete algorithm, but it would help knowing more about the CPU temperature thresholds used and about other sensors that may greatly impact the fan speeds (just guessing that things such as GPU utilisations and hard drive temperatures have a strong influence). We paid a lot for a laptop where the fan control algorithm mis-behaves, I think knowing some details that might help us solving the problem is a kind of our right.
6) What is the temperature of the testing room where the engineering group works? Is the room icy cold?
7) Why are you testing the laptop only in 20 minutes batches. It doesn't make much sense, since people experience the fan stuck behaviour after hours of use.
8) The laptop runs icy cold and yet the fan goes on like crazy. Why the engineering group did never consider adjusting the thresholds used by the fan control algorithm, such that browsing the web doesn't trigger an increased fan speed?
At the end, I'd love to note how having an open source BIOS will help. I look forward to the day where the BIOS will be available to users putting an end to such kind of issues.