Regarding the haste cap to aim for people who cannot reach 2005 with their current gear: I was playing a bit on TreeCalcs the other day, and it seems to me that the 1779 cap (6th rejuv tick with NG) is quite valuable. There is a combination of factors here:
- At 1600 haste (plus or minus a few points), the inital rejuv tick from GotEM is increased by around 600 (in my current gear), which comes down to a 2% increase in the total rejuv amount (probably even a bit more because the initial tick tends to do less overheal, but assuming 3/3 GotEM of course). This is coming from the strange way this initial rejuv tick is computed with respect to haste. I remember this was described by Hamlet in an early cataclysm thread but I guess the thread has been removed now. Anyway, I have tested and confirmed a few days ago on live that it is still behaving in the same way in 4.2.
- It is easy to maintain a NG uptime of more than 20%. So the 6th rejuv tick during NG provides on average a 4% increase on the total rejuv healing amount.
- At 1594, you also get the 10th WG/efflo tick under NG.
The main argument for staying at 916 haste is that rejuv does not benefit from haste, but combining only the first two factors above, we see that rejuv gains around 6% increase by going from 916 to 1779 haste, which is quite close to what other heals gain in cast time reduction (863 haste rating = 6.7% additional haste). It is even slightly
better than what the same amount of mastery would provide: (863 mastery rating = 6% harmony, but worth less since it is additive with other bonuses).
This is average healing only, but NG is more versatile and can be kept for periods of intense healing if needed which is yet another advantage for reaching this cap.
Of course, 2005 is not far from this and is an even better cap to aim for, but for people who cannot reach it in their current gear, I think is is actually better to go for 1780 than reforging down to 916. And this should be the case for both 10m and 25m.