I'd be interested to see complete, updated-for-2.4 calculations for lacerate as well. The estimate posted by dukes is a good first order approximation, but there are additional factors to consider if one wants a precise point where swipe threat overtakes lacerate. Honestly, though, there is probably some intermediate range where keeping up a five stack of lacerate and swiping in between is actually best for maximizing threat.
Besides the question of RED/PED in meta socket, lacerate has additional value when it misses/is dodged/is parried because the dot will continue to tick for some threat as well as while using other abilities such as Mangle/FF/Demo Roar if applicalbe (yes, it's not huge, but since it now ticks for 400ish that's 40 threat from damage per GCD assuming 20% modifier). Lacerate also applies nearly full threat if blocked, though swipe now also does considerable threat when blocked because of 2.4 block changes.
Lacerate further benefits in situations where rage is not effectively unlimited. Besides enjoying a base cost of two rage less, lacerate's rage cost is reduced when it is avoided while swipe will still cost the full 15 rage. I sometimes find a few seconds tanking Azgalor or Kalecgos where I end up rage-starved because of a string of avoided attacks, not to mention the need to be as efficient as possible on a boss like Bloodboil.
Threat considerations aside, I don't think the value of a 130+ dps dot should be discounted. Even if swipe generates slightly more threat, keeping lacerate up between swipes in order to maximize DPS contribution to the raid seems valuable. I frequently "roll" a lacerate while using swipe as my primary non-mangle GCD move, usually refreshing lacerate at around 6s in case of bad miss string or needing to mangle.
Since I'm bored now, here's a first-order approximation similar to dukes' old one based on my normal lacerate stats in tanking gear (2t6, 2s3, heroic legs, no RED). I assume that lacerate changes much more slowly with gear than swipe. I ignore misses/dodges/parries/blocks and dots ticking while not casting lacerate. I'll assume the same 30% crit that dukes used.
Lacerate threat = (DoT threat during one GCD + Average application damage) * (Lacerate damage modifier) + Lacerate inherent threat
Lacerate threat = ((405*.5) + 50*(.7+2.2*.3)) * .2 + 285
Lacerate threat = 335.5
So to compete, swipe damage would need to be solve 335.5 = (dmg*(.7+2.2*.3))
so dmg = 246.7
Having actually done this calculation, I'm not certain how dukes arrived at 335 with the old lacerate. It seems he was factoring in the entire threat value of a single lacerate stack over its lifetime into a each lacerate; I can't think of any situation in which this is actually the value of a lacerate except when applying a single lacerate and not stacking it. If you modify lacerate to be applied exactly every 15 seconds (also not really realistic but easy to approximate and an upper bound on the value of rolling lacerates), adding in the entire lacerate dot rather than the half tick you get during a GCD gives lacerate's threat value to be (405*5 + 50*(.7+2.2*.3))*.2+285 = 703.6 (someone might want to double check me on that). Unless someone can show me a tanking gear set and boss for which swipe does over 700 damage on average (468 non-crit with 40% crit and RED, though with that kind of AP I bet lacerate ticks for a lot more), it seems like rolling lacerates is probably a pretty good idea. So according to my approximation, from a threat perspective one should spam lacerate until swipe does around 250 damage, at which point one should roll lacerate between swipe spam. That system shouldn't be replaced with pure swipe spam until swipes hit for almost 500 (and as noted if I had anywhere near that much AP I'd be re-doing the calculation).
I apologize if anything I said here is repeated, worthless, or otherwise idiotic; I'm pretty new to these boards so go easy on me.
