One other interpretation is that the short term fuel trim is the ratio of 14.64 to Lambse at any given time when in closed loop. I.e. if Lambse is 13.31, your short term fuel trim is 14.64/13.31 = 1.1, i.e. the EEC has added 10% more fuel to acheive stoich (I may have this backwards, i.e. Lambse/14.64 to give lower numbers for more fuel added). This trim is only applied to the fueling at hand for the purposes of approaching stoich, i.e. short term, in both derivation and use. The EEC remembers these corrections (or their averages) at given conditions and uses them, over time, to update the long term fuel trim table (KAMFR). The KAMFR table is recorded and applied according to the axes of the KAMFR table (RPM and load?), and is kept for a long time (hence long term fuel trims). This process of updating the KAMFR table is otherwise known as adaptive fuel control, and turning off adaptive will force KAMFR to stay at 1.000, and you have effectively disabled the development of long term fuel trims. If adaptive is enabled, then once KAMFR is properly updated, the short term fuel trims will be very close to 1, i.e. Lambse will be close to 14.64 in closed loop, while KAMFR will be something other than 1, with different values based on the position in the table based on load and RPM as developed by the EEC's observation results of closed loop. In other words, closed loop is used to calculate short term fuel trims required to acheive stoich, and adaptive fuel control takes these short term fuel trims, and by also observing the RPM and load they were obtained at (with averaging), develops the long term fuel trims, otherwise known as KAMFR.
I still haven't decided which to believe. I like the latter, since everything is well defined However, it presumes that KAMFR is applied TO the Lambse value to determine the pulsewidth (as opposed to being used to determine Lambse). People have also observed that if adaptive is disabled (and KAMFR is fixed at 1), that their WOT tunes stay more stable, i.e. don't get rich or lean over time, suggesting that 'long term fuel trims' have been disabled.
If, perhaps, Lambse was expressed as 1 instead of 14.64, then it would be more intuitive - you would see two numbers that vary up or down around the value of 1.000 - i.e. Lambse would run up and down during closed loop - and it would be easier to call that Lambse value near 1.000 a ' short term fuel trim', while KAMFR would vary around 1.000, according to the axes of the KAMFR table, and be called 'long term fuel trim'.
Comments? I've never heard anyone definitively state what the correct method is. What does the Ford documentation say?