We certainly look at ToRParse for data, but like all other data sets it is limited in what it shows and what context influences it. For example, if I were to sort the top DPS value on Kephiss in EC 8-man mode, I would immediately see that a Mercenary is listed at the top of the charts. However, if you dig into the charts, you can see that part of that number is based on a bunch of trash kills that have influenced the final result. When we detach the source (the player) from the log, it makes it much more difficult to look at context. All of that being said, we really want to do better about collecting raw logs directly from the player, with the hope that will really help us properly react to the feedback we receive.
Actually, the problem with that log (I assume you mean Cay'liy) is not the Trandoshan adds, but rather the fact that the combat log and/or parser bugged, resulting in an aborted parse. The *only* part of that fight which is represented in that combat log is the trandoshans, which is obviously misleading. The second-highest DPS is a scoundrel (and a guild mate of mine), who has a similarly abbreviated parse due to Vanish. The highest *legitimate* parse on that leaderboard is, to no one's surprise, a Powertech. Continuing down the line, we have a Maurader and a Commando, then a Vanguard, a Sniper, a Gunslinger, etc…
A better fight for this is the Dread Guard in Asation, where the 8 man HM charts read like a list of FOTM classes.
I'm not quite as skeptical as most; I do think that the class balance in this game is at least within spitting distance, all things considered. However, it could be a lot tighter. My aforementioned scoundrel guildy rolled a vanguard just for kicks. He decided to main swap when he discovered that his vanguard was doing 250 DPS more than his scoundrel…where the vanguard was in 58s and 61s and his scoundrel was in min-maxed 63s. That should never happen.
If nothing else, torparse has a phenomenally good repository of combat logs from the game's top tier PvE players. While I would be very surprised if Bioware were actually using torparse's leaderboards to make decisions, they absolutely should be data-mining the repository to augment their own analytics. This sort of thing is very important, I think. When the community as a whole has come to a (essentially unchallenged) conclusion about the state of a particular class, Bioware should pay attention. When that same community can provide a mountain of empirical evidence to back up this opinion, Bioware should act. All we're asking is that the class balance should stem from the data. Right now, the data isn't looking very positive for certain classes.