It looks to me that the dates are decimal, not hex. I think they're trying to say it's encoded, not human readable.
From Python:
time.struct_time(tm_year=2005, tm_mon=8, tm_mday=11, tm_hour=18, tm_min=13, tm_sec=21, tm_wday=3, tm_yday=223, tm_isdst=0)
So that's year 2005, 11th August. Could that be when the firmware was first written? More likely, there is a different epoch involved (Python seems to use the Unix 1970/Jan/01 epoch).
"Programming" time may not be when the firmware was updated; it could be when the firmware was written, or compiled, or indeed when updated to the ECU, or perhaps other interpretations.
A quick Google search didn't bring up anything hopeful.