66i Expedition Mode recorded elevation in ver. 9.0

Has anyone checked for elevation errors in recorded tracks made when the 66i version 9.0 is occasionally resumed and then returned to expedition mode? 

Through several past firmware versions I got erratic elevation recordings when hiking through mountains for several hours over several miles, but I have not tested the last two or three versions. To be precise, the error I refer to was only obvious in recorded elevation.  There was no corresponding erratic error in horizontal position.  And the erratic elevation track recordings only occurred in Expedition Mode.  The errors occurred every time I used the 66i on a long hike through mountains in Expedition Mode.  I am not referring to the usual outlier readings that apply to only one or two track points and where subsequent track points return to reasonable measurements.  I am talking about sudden errors of 50 to 1000 feet that occur between track points a couple of minutes apart and then persists through all subsequent track points.  I believe the elevation tracking was set to calibrate once at startup, but I'm not certain.  The problem may have resulted when the 66i was in continuous recalibration mode.  Also, I'm not talking about satellite tracking.  I'm talking about the track points recorded internally.  The errors could be related to times when I resumed the 66i for a few minutes to check something and then returned it to Expedition Mode.

I haven't checked for the error in the last two or three firmware versions.  Once I realized I could not trust elevation in tracks recorded while in Expedition Mode, I stopped using Expedition Mode whenever I needed a reasonable elevation track.  But I would like to.  That's why I'm asking here.  Expedition Mode is a wonderful and amazing feature when you have to carry all the battery power you will need on a long hike over several days.  So it would be great if someone has the time to record some long tracks over several miles in Expedition Mode when the 66i is resumed and returned to Expedition Mode a few times during the recording.

Usually, the errors I encountered resulted when I was trying to keep a single track over a one-or-two-week hike where I was moving every day.  But on several occasions for experimental reasons when I did not need an accurate elevation track, I checked on a shorter hike over six or seven hours and got the same results.