Just a few notes on these:
1. SB potency wouldn't necessarily be overpowered, especially at such a low cleave potency, even if its main potency were left the same. Moreover, you're going to lose out more than you're thinking here by shifting 25 potency from SB to RA to compensate, just because RA has no enmity modifier. Definitely a fan of all cleave-looking abilities cleaving, though, and SB seems the most popular choice among GLD abilities of late.
2. One of things that's interesting about GLD/PLD oGCDs is that they all contribute identical single-target potency per second at the moment (150 per 15, 250 per 25, 300 per 30), if used ideally (full health SW, Swipe used instantly upon CD refresh). CoS is already the strongest oGCD, even in single target, just because it has none of the limitations of the other two.
3. If it dealt damage, it would also break CCs.
1. Would be nice. One of my favorite suggestions was to just place all magics somewhere along a spectrum between purely aetherial (Unmend, Bio, Fire) and purely manifest (Stone, e.g. magically flinging a giant, very much physical, boulder).
2. That would be pretty OP at that point unless average block chance were reduced to compensate, which would then just give you a feeling of being half-assed whenever Bulwark is on CD, especially if blocking were more important to GLDs/PLDs.
3. Well, it is the same TP cost and potency, basically, as One-Ilm Punch. But, it already comes with a stun, which is practically taking up to 6 seconds off all its buffs anyways.
1. /shrug
2. Can't deny that.
3. Element aside, I'm not sure why a 10%HP shield would feel unlike a Knight or Paladin.
4. Just beware that when you lengthen any duration, it messes with the classes that already use the ability, e.g. Monks. I think it would be wiser just to raise the DoT's per tick potency from 20 to 25, increasing the total damage from 220 to 250 such that it's still TP-inefficient, but at least over RA-average potency.
1. It makes no sense for a weaponskill to generate flat TP. At that point you should just reduce its cost instead. The same could be said for Shield Swipe if it was made a viable weaponskill at below-average TP cost. Aside from that though, if you allowed blocking for all magic attacks, Swipe still wouldn't balance PLD TP-generation specifically around physical fights, which would otherwise be the main downside of this.
1. Great.
2. Our Oath-swaps are still pretty good as is. We also have the option at least of dropping Shield Oath just before finishing our combo and then popping Sword Oath before the next (well, next-next) AA, which largely mitigates the combo-cost. The GCD loss itself may be a bigger factor in most cases. Just throwing info/ideas out there.
3. Could also limit it to weaponskills, reducing its penalty to effectively 13% or so.
4. Makes sense, given the first, but would come with costs in certain rare situations (e.g. where splitting damage between two tanks from a mixed magic/physical boss would have been preferable).