Although a player's "rotation" will vary from phase to phase and even within a phase depending on dynamic haste conditions, the impacts of moving and interruptions, and RNG, it still helps to have an ideal rotation to shoot for under each scenario - e.g, I want to do this during CA and that during Standard unhasted. With looking at the various haste ranges or requirements for each scenario and ideal rotation, you can determine a rough haste range or amount that best fits you. Furthermore, you need to be at least aware enough of the haste requirements so that you can balance both a high DPS rotation with focus consumption.

For instance, lets look at the following case:
- Use AS as the focus dump outside of CA phase and when not under dynamic haste effect; otherwise, use AI.
- Has T11 4-set
- Has the SS/AS/CS glyph combination

Since the Standard phase unhasted is the portion of the fight that you should normally be in the longest, I like to start with it. The recommendation in the OP is a CS-ASx2-SSx5 cycle that has a high-limit of about 15.12% for a tight CS rotation (not including latency and reaction times) on cycles without an MMM AI proc but to have about 0.2s deadtime on cycles with a MMM AI proc. It indicates that you only need about 10.52% to have a tight rotation on cycles with an MMM AI proc, but that you will have about 0.25s delay in your CS cast on cycles without the MMM AI proc. So ideally if acquiring haste had no cost, to maximize the benefit of the CS glyph during this situation, you want somewhere between 10.52% and 15.12% haste. However, with acquiring haste having a cost in crit and mastery and how much that cost is depends on your overall ilevel of gear, the actual maximum DPS point varies from player to player. They key though is that the ideal range is between 10.52% and 15.12% with realizing that even less haste may be okay with the penalty of delaying your CS casts more.

When hasted during the Standard phase, the best rotation appears to be CS-AIx2-SSx5. This only requires 6% haste from gear to be both tight and sustainable, even on cycles with an AI MMM proc. Since this haste requirement is less than that for the Standard phase unhasted, it is already supportable by your choice for the Standard phase unhasted and skews the maximum overall DPS case to be at closer to the lower end of the range.

What is probably the second most important maximum DPS phase to the Standard phase unhasted is the CA phase during RF. It has short duration (only 30s), but does by far our highest DPS, especially with trinket procs and racials. The recommendation is CS-AIx3-SSx7. This rotation is supported by the recommend haste range and results in a cycle time of just over 12s with the amount over depending on where in the haste range you are. It provides about 2.6 to 2.9s of leeway in refreshing SrS with the recommended haste range. Focus is not really a concern with this rotation within the recommended haste range. Where you are in the haste range does not have a large impact since the cycle time only varies by less than 0.2s. A little bit less haste than the low end of the range is probably still fine.

When unhasted during the CA phase, a CS-AIx2-SSx7 rotation is recommended (same rotation as during CA RF but with one less AI). The recently added table in the OP illustrates that with the recommended haste range that each cycle will have at least 1 to 2s leeway to refresh SrS and even more time on cycles with a MMM AI proc. This cycle is also recommended during Bloodlust in the CA phase but with the extra haste resulting in a shorter CS cycle (with more leeway in refreshing SrS) and better focus balancing.

I have not gotten around to analyzing the Termination phase since its lasts only about 5% of the fight length and many people are not talented for it. It is low priority and probably not worth changing your Standard phase rotation in many cases. Although, I do recognize that in some situations that DPS could be gained by maximizing your rotation for the additional focus regen from SS casts.

I still need to analyze the KS phase as well, but I am not too concerned with it since it should have similar outcomes as that for replacing shots with AI MMM procs, except now there are more occurences per cycle.

Anyway, the rotation analysis definitely indicates that although an exact haste number is not recommended for which to shoot, that a range of haste from gear between about 10.5% to 15% is ideal with where you fall in that range being dependent on what your gear can ideally support and which fight phases you want to maximize for your personal preference.

Similar analyses can be performed for other situations, but here I just provided the information for the case I am currently following. For my gear, just getting over the 10.5% haste from gear is about the amount that my gear can support.

I believe you are not grasping the difference between Priority, Rotation, and Shot Cycle with flexible order. Using simcraft with a priority based on focus levels and buff time remaining comes close to the shot cycle, but its not a sim of the shot cycle itself.

If you have some knowledge in basic Discrete Math, think of the "Shot Cycle with flexible order" as a Combination (math concept) of elements in which, for this particular case, SSx2 must exist at least once and CS must be used at the beginning/end of the cycle.

I do not know whether this answers your question or not. Unglyphed RF is 40% haste, and glyphed is 50%. If you want to know how much haste that is averaged out over a fight, it depends on the fight length and how many times you used RF. Using a 5 min fight as an example, you should be able to use RF 4 times for 1 min of the 5 min duration, or 20% of the time. Hence, the average effect of RF over the fight is 8% unglyphed and 10% glyphed.

However, I would not suggest using the average numbers in any analysis since it is misleading. Using the average number is not too bad when looking at autoshot and WQ procs from them, but it is when looking at specials and the number of them that can be performed as well as the impacts to your CS cycle time. In addition, 8% additional haste may not be enough to make using AI worthwhile instead of AS, but 40% haste definitely is.

The trinket you list appears to provide about 12% (normal) and 13.5% (heroic) haste for 15s every minute. Thus, the average values appear to be 3% and 3.375%. But once again, using the average value is misleading. An additional 3 to 3.375% haste is not necessarily enough to make it worthwhile using AI instead of AS as your focus dump, but 12% or 13.5% is. At 12% from the trinket and 10.5% haste from gear, AI casts are under 1.8s when the trinket procs, making it worthwhile to cast if able too.

Hence, especially with the base agility, this trinket has some possibilities but will have to be weighed against the other trinkets once we see the full list.

As far as how much DPS can be gained by using the trinket, there is no good answer. It not only varies from player to player but varies for a player from fight phase to fight phase, with shot selection, and depending on how well they are stacking other procs with it and utilizing the haste.

Since you can control when the effect occurs when off its ICD by going below 20% focus, you can use it during the CA phase to reduce your AI cast time. However, even in this situation the benefit varies depending on when you use it and how long your CA phase is. If your CA phase is only 30s, then I would stack it on top of RF for very fast AIs. If your CA phase is longer than 30s, then I would use it after RF to not only haste more AIs but to also haste your SSs so that you will have both more focus and time to cast additional AIs in the phase.

During the Standard phase, I wouldn't stack it with RF since using it with RF doesn't affect your SS casts much if at all. I would use it outside of RF to increase hasted AI cast times and so it affects SS cast time as well.

Thanks for correcting it. I've put a cautionary noted bolded on top of my post. Maybe you could a short note of this analysis to the op.

Oh and silly me, I made the most common scientific mistake of all. I did see that extra 100 factor, but thought it must be a mistake due to percetange/decimal change at the start of my calculations and simply removed it. You know, because the numbers were so large and that just can't be...