In light of the recent announcements and Cal's good threads on PvP and WvW balance, I wanted to take a minute to express a few thoughts/concerns on balance in the bigger picture -
Most people seem to be excited by the idea of doing more skill splitting between PvE & PvP, but I'm not sure I share this. It feels a bit like band-aid fixing things that are facets of larger systemic issues. I think it's important to keep a holistic perspective on combat balance across all three game modes and avoid letting them become too siloed. Many things get missed when there's a hyperfixation on balance in one mode or another. There are a few examples I can think of right off-hand:
Historically, balance changes have been more concerned with PvP (even though PvP balance isn't particularly good). This has a few effects, but one obvious one is the assumption of shorter engagement times, which diminishes the role of the #1 skill and enhances the role of cooldowns in an encounter. One of the big "misses" here is the relative balancing of auto-attacks, which carry a lot more weight in PvE than they do in PvP. Since the auto-attack on a weapon sets the baseline damage pressure of a weapon in a longer engagement, a weapon with an undertuned #1 skill will absolutely underperform in PvE while it doesn't necessarily in PvP. This is something that at times has gone years seemingly without being noticed or corrected (such as Thief's P/P or Ranger's Longbow) and in some cases is still a problem in the game (such as Elementalist staff). The issue is there doesn't seem to be any coherent design philosophy around how this is balanced and managed. Maybe I'm wrong, but I'd be surprised.
Other examples are the efficacy of control effects & stealth which are also balanced around PvP and therefore the assumption of short engagements with quick deaths/quick resets. This leaves most control effects as well as Stealth feeling underpowered in PvE. Stealth should be a fun strategic utility in PvE when instead it's mostly just used to get big damage on an opening attack. Shouldn't there just be a general design rule that control effects and Stealth last x% longer in PvE than they do in PvP to accommodate the difference between mobs & players, and wouldn't that help reduce the need to try to balance individual skills between PvE and PvP without touching their damage components? After all, shouldn't mob statistics be modeled after player statistics so skills can work in the same way for both?
Secondly, a huge portion of the unhealthy power creep in the game revolves around two things - a.) the over-proliferation of boons, and b.) the mechanical changes made to conditions since launch. In my opinion, these are not PvP issues but are game-wide issues. Condition damage scaling is too high with durations that are too long, so there's a reliance on overpowered mechanics to combat it. For example, most condi-clearing skills and traits are overtuned as a way of dealing with high condi damage and Resistance negates all condi damage instead of doing something sensible like reducing it by 50%. In PvP, this forces players to build around being able to deal with incoming conditions. In PvE, it means that too much condi damage goes to waste, making it generally inferior to Power builds rather than simply a bit more specialized like it should be.
It really feels like most of need for skill splitting between PvE and PvP is to deal with issues that are more foundational and needing to be corrected either with mob mechanics/AI or with generalized differences such as the control one above.