A good rule of thumb I use to get around the issue of valuing hit after reaching hit cap is to convert hit to spell power at the same ratio as the amount of hit on a yellow gem vs. the amount of spell power on red gem.

One quick question: Is the [Extract of Necromantic Power] still considered one of the worse trinkets around, even for Affliction?

What if you are destro?

Yeah, it is. Even back when Affliction could put up Immolate and Siphon Life was a separate spell, it wasn't too hot. The 15 second internal cooldown was good, but the damage was weak and it didn't proc too often.

For two easy to get, effective trinkets, go for Sundial + Dying Curse and build from there.

Has anyone done the math to see if [Dying Curse] yields a higher net dps over [Living Flame] because of the shorter internal cooldown, assuming one would be at or over hit cap using either?

I think you're still missing the point. You value hit as 3.27 DPS UNTIL YOU ARE CAPPED. You end up with too much hit by taking too many hit items not by weighing hit too strongly. If that isn't obvious to you then I give up.

Dying Curse:

10(Duration)/45(Minimum CD) = .222222
.222222 * 765 = 170 effective SP increase

Living Flame:
20 / 120 = .166666
.166666 * 505 = 84 effective SP increase

This assumes Dying Curse procs every time it's CD is up, which it doesn't. It only has a 15% proc rate. So, lets use some VERY rough probability:

100% / 15 = 6.6667 <---- This is the number of casts that (roughly) need to be thrown to get a proc.

With my haste, Incinerate is 1.7 seconds.

6.6667 * 1.7 = 12 seconds of casting including a small amount of lag

Dying Curse:

10(Duration)/57(rough CD) = .175438
.175438 * 765 = 134 effective SP increase

Dying curse is better, assuming you're hit capped either way.

Here's the big wrench though. Living Flame is a use trinket. You know you can use it on the most effective part of your rotation, where as the Dying Curse can proc at the worst time, say, right before you have to move, or there's a phase change and 0 DPS is being done. For destro, lets say, you can get 2 immolates, and a curse of doom into every trinket cool down of the living flame. This provides a huge boost to conflag too. Or, you could save it for the burn down of XT's heart, etc, etc. In conclusion, I'd weigh the Dying Curse more on tank n spank fights, where I'd use the Living Flame on burn down, high movement, and presicion fights ... aka Ulduar.

Not to mention that while on paper Dying Curse has effectively 50 more spell power, you could potentially make that up by switching out other pieces of gear due to needing 36 less hit.

Hit as x dps until you are capped is a more reliable indicator.

It is entirely possible that the optimal gearset for you can turn out to be 1-2 points shy of your desired hit cap, purely because you may well have to choose between being ~2 points under hit cap or ~4 points over. In this case, the 'lost' dps from being below hit cap can be noticably less than the 'lost' dps by sacrificing stats to get those last half dozen points of hit. This is something you simply would not pick up if you always value hit at 1dps, or some other convenient value.

We may not be talking about life changing numbers here, but when dealing with optimisation it's important to be as accurate as possible.

This is absolutely 100% true for a target dummy. However, I would argue that any chance at all of missing a single spell on the heart during a hard-mode XT-002 Deconstructor try, for example, would make a gear set 1 or 2 hit rating below the cap less than optimal. It all depends on your definition of "optimal."

I don't believe that any trinkets share a CD. I have dying curse/sundial and they both proc at the same time all the time.