While 'fooling' keeps the stock ECU from throwing a code, if stored data can actually be read or printed out it may well be that the data itself would show the actuals, meaning that a printout in graph form might demonstrate different durations, fueling-hp-etc and the existance of alternative programming or at least alternative operation.
This information may be obtainable in the overall picture of analyzing performance or diagnostics. It just might not be part of a normal testing procedure that is run in the overall checking of a truck when connected to equipment to do flashes, or ECM/PCM even ABS adjustments.
While the more noticable physical signs (velcro, EGT probes, gauges,etc) are clues regarding 'boxes' possibly existing or an assumption that a 'box' exists, a programmers is a whold different ball game.
Ultimately codes disapear with x-number key starts because your reference is the idiot light finally turning off, but do they really? Is there a recording process that requires them being physically removed? Seems to me that must be the case as the Smarty provides that ability no differently than the Star Scan in accessing this ROM area and 'cleaning' the information. As I understand it, and I could be wrong, the Smarty programming does replace the ECM programming. By 'returning to stock' it returns the factory programming, the firmware so to speak.
Having the separation of programming vrs a secondary catche of data is not by any means some new proposition, in fact, that is pretty much the satandard method of operation regarding ECM's and keeping them safe from some alternate source of data corruption. Any high magnetic exposure will erase your firmware on your ECM, the RFI from a CB radio can run interference or affect your ECM.
It would be equally conceivable that the 'dents' also disapear in/over time with key starts. I doubt that the ECM is capable of keeping data for the entire life of the truck in a ROM, but it may keep sufficient information in a ROM for x-number hours of operation, for the purpose of diagnostics.
Again, while 'fooling' takes place as to not set off codes, some other form of information for the purpose of diagnostics may be there in the last x-number hours of operation/running of the truck. That data also might surfice to actually be the point of detection or difference over what would be expecte to be normal.
We all know that a ton of data is retrievable with regard to operation, that an insuracne company can be provided with the precise moment in time that an airbag was set off, at the same moment in time they can also know what your speed/rate of travel was, rpm's, possibly the gear your transmission was in, the exact timeframe it took your reflexes to apply the brakes, and furthermore they also can know your exact direction of travel via the data from the overhead information being streamed in/out of your trucks computer systems. This simple information requires running software on your truck to extract data, plain and simple, all data can be extracted from some point in time. If the 'fooling' also serves to change the performance mapping, it would still be hard to change how that mapping sees duration times, etc differently because running a parallel shell would not be able to compensate for duration.
CD