symmetrie of tenses

["Alfred W. Tueting (Tüting)" <Ti@fa-kuan.muc.de>]

Events take place in space and time - most significantly in time
(e.g. mi pensi) in a linear extension into two directions ba and pu 
(in relation to the subject /mi/ or to the event /le nu mi pensi/
respectively). I regard these two linear extensions as two more or 
less extended/defined time intervals (the two end 'points' being /le
mi ca/ and /mi (ba/pu) co'i pensi/ or /le nu mi pensi/ and a 
point far far in the future (ba) or in the past (put). That far, this
is obviously symmetrical.
Looking at the event /mi pensi/ itself and its contour, imbedded in
time, I'd like to see several subintervals divided from each other 
by 'points':
1) One reaching up from far far past (pu) to a 'point' (or at least
tiny interval) of time called /pu'o/ *immediately adjacent* to  
the beginning of the event /mi co'a pensi/,
2) second (or third one, if /pu'o/ is regarded as an interval itself)
represented by /ca'o/, reaching from the point /co'a/ to that of 
/mi co'u pensi/,
3) the next one (roughly speaking!) from the point /co'u/ to a point
far far in the future (ba).

Now my problem and question: If I understand the Book right, then
/pu'o/ is regarded as a point or very short time interval *
immediately adjacent to /co'a/, isn't it? (p. 228, 10.1) "mi pu'o
damba I [inchoative] fight I'm on the verge of fighting"). I'm now 
pondering whether or not /ba'o/ is symmetrical to /pu'o/ in this
regard, i.e. also a point or very short interval of time, adjacent to 
/co'u/ on the other side? Or, different from it, a somewhat undefined
space of time extending into /ba/, that I really wouldn't be 
very fond of!
I'd like to translate e.g.:
mi pu pu'o damba... I was *on the verge of* fighting, (when ...) and 
symmetrically
mi pu ba'o citka le mi sanmi... I *just* had finished my meal, (when
...),
but I really doubt whether this is possible.

Any comments?

.aulun.