There's no point in comparing their values over 4 GCDs when GB requires nearly 10 GCDs (at 2.5, or exactly 10 at 2.4) to deal its full damage. It should be considered in the window of its actual rotation.
Let's look at the potency dealt over 2 GBs, 48 seconds, using the 2.4 GCD, starting from the application of the first GB and ending in the application of a third, under the two situations.
Clip: [500]+150+200+340+150+200+340+150+230+500 :: +150+200+340+150+200+340+150+230+500+150+200>> 500 + (690*4) + (880*2) + 150 + 200 >> 5,420Swipe: [540]+210+150+200+340+340+150+200+340+150+230+540 :: +210+150+200+340+150+200+340+150+230+540>> 540 + (690*4) + (920*2) + (210*2) >> 5,560That's a 120 lead in potency, (not so) coincidentally the exact sum of simply having not clipped off 1 tick of each GB.
Now, if you blame my window of observation for these results, consider:
With the GCD thereafter, Clip will add 340 and Swipe will only add 150 (5760 vs. 5710, clip's lead), but...
(but that is not a rotation; ultimately you will still only be able to use two RA combos. You may reach your later instances of RA sooner, but you won't have more, rotationally, than the Swipe rotation.)
Add a second GCD and that's 150 to Clip and 200 to Swipe, and they're equal (5910).
Add a third GCD and again Swipe leads, 200 vs. 340 (6110 vs. 6250).
And it will continue to vary with the latest nuke, which Clip will reach faster, until Swipe's higher rotational strength exceeds the difference between the start and end of combo (5 RAs). But if we're looking at rotations here, then... look at the full rotation (GB window).
Again, this is dependent on, explicitly, a 2.4s GCD as on my own PLD. At 2.5s you are actually looking at the possibility of wasting a tick vs. missing an opportunity for a tick, where there isn't nearly such a clearly preferable rotation. But if look at GB as it is, rather than as a spammable weaponskill, with the GCD you quoted, using Swipe once per GB is quite simply a better rotation (unless you somehow know the exact second at which a boss will jump, would have been able to end with a RA instead of an SB if you hadn't Swiped, and if said boss will jump in less than 4 completed GBs).
Again, it has nothing to do with the alternative potencies for that button push. It has to do with the rotation that will be caused by pushing that button.
There's always a move that is stronger than a clip. It's the same DoT, without the clip. Unless avoiding the clip prevents you from using a better timed move down the line (again, a matter of... avoiding clipping), you should always avoid the clip. You want to get the most out of every buff and debuff. If it requires that I use a individually inferior filler to get, in total, more out of my overall rotation, I will do so.
A Monk is at Raptor form with 6s left on Demolish, 8s left on DK, 10s left on Twin, and Fracture not currently up or with less than 4s remaining. Is he going to clip or Fracture? Or would you rather sac 9.6% of your Demo and 13.3% of your DK and Twin, in order to get to the same exact loss of potential potency 1 GCD earlier as the next Demo nears its end?
DRG's rotation is unique in that, frankly, it has two GCD plateaus, one of which fits the needs of Geirskogul/WT/F&C, and another pre-BotD rotation that fits its original ability windows (CT/FT combos, HT, Phb). The 3.x rotation must sac a full CT in order to hit HT and Phb at their perfect times and simply because the only alternative to using another CT 5-7 seconds early is to apply it 5 to 3 seconds late. A pre-F&C/WT DRG can fit 2 Full Thrusts, Heavy Thrust, and Phlebotomize into each CT, losslessly, or 3 Full Thrusts and losing HT's buff on its reapplication only, using a 2.33-2.35 GCD, such that each buff/debuff duration is reapplied with less than a half-second to spare. To say that a job that has shattered its original rotation in order to uniquely support a 600p/min oGCD specific to its job is a prime example for all others is ridiculous. You have abilities, and you seek to get the most out of them, for the greatest overall dps... That's the prime rule--greatest overall potency, not 'most potency in a given GCD'. (How do I not get myself into this bind? > what's my best option, now that I'm in an inferior position?)
Food for thought: by increasing Skill Speed in a 3.x rotation and dropping Phlebotomize, you can fit an extra FT (and therefore WT/ into each CT, allowing an extra Geirskogul per minute, a free, no-compromise, 200 potency or 3.33 pps. In the normal rotation, Phb is maintained perfectly, providing 410 every 24s, or 17 pps, but at the cost of a GCD and other rotational opportunities. Maintaining HT into the next HT gives another .85 pps. An extra FT+1 per CT gives an extra 1000 potency per 30s, or 33.3 pps. Getting another 2-3 ticks off CT gives it another 35-70 potency or 1.7-2.3 pps. If Skill Speed were weighted equally with Critical strike, which then would you pick?