His parse is for 3436.56. When you pull it back to the "enter combat" and set the limit to 5 minutes it comes out to 3358.31. A changed of 78.25 dps. In your example You give him an extra minute on the parse. These are just details, however there is a point to it (I am not picking them out just to argue). Saboteur (with scaterbombs) is not necessarily a sustained or "build up" spec. The damage comes strictly from the 20 second CD on HI/CE. It becomes a burst spec with dots that keep some damage going. Cut a log before they go off, the dps falls. Cut a log after it has been used dps rises. The damage the spec does can be manipulated on paper for a desired result.
Different specs require a different amount of set up to do damage. If we are talking about sustained, how about I run a Lethality/Dirty Fighting parse and cut it to only start when the first Cull/Wounding Shots starts (Eliminating the dots in the beginning for set up). This would allow be to get rid of the "ramp up time" and focus more on sustained because Lethality/Dirty Fighting focus's on the proper use of Cull/Wounding Shots. This method is not true to the spec considering what it actually does.
If you analyze a Watchmen Sentinels parse, you would have to cur everything before the 6 stacks of Juyo and 4 stacks of Merciless Slash. So, if I where parsing on my sentinel (by your rules) I would initiate combat --> build focus and centering (saving Master Strike and Zen) until I get my stacks of both buffs --> I would then begin doing damage with all relics and adrenals going with Zen and Inspiration at my. I would then cut my log in the begining to eliminate all of the build up of focus, centering, Juyo and Merc Slash. The log would run for 5 minutes following. This would give me a better idea of what sustained damage I would be doing.
To give a representation of this, I got on my LvL 50 sentinel and went to fleet (yes I had to deal with that nasty lag) and parsed on the LvL 10 training dummy. I had to use that one to ensure I hit with my abilities since the rest are LvL 55 (My gear was a full 63 BiS build from pre 2.0....it included accuarcy stats). Here are my results:
Not Clipped: (2433.68) http://www.torparse.com/a/244058/2/0/Damage+Dealt
End Clipped: (2421.23) http://www.torparse.com/a/244058/tim...0/Damage+Dealt
Beginning and End Clipped: (2477.75) http://www.torparse.com/a/244058/tim...0/Damage+Dealt
Now I ask, which log gives me my sustained damage, my damage threshold and my realistic damage?
The moral of the story here is that importing logs that were created under a different rule set skews the data in favor of trimming the beginning of each log. It sets those logs (unfairly) above the rest for no reason but to measure an e-peen by presenting data in a different fashion.
On the note of measuring fairness and actual thresholds on a global scale, I would put forth the following rule set:
Let me get right on that for you
been trying to get a nice MM parse, but I keep getting thrown out of combat