The primary assertion comes from the fact that Feinting in Combat is generally effective when the player has maxed out their Bluff ranks, but in the handful of encounters where the target has maxed out (or even just high) skill ranks in Sense Motive, it's tipped greatly in favor of the target. This seems a bit unfair to the character investing in the Feinting in Combat tactic.
For example, with a 5th level Rogue with a 10 Charisma, Bluff maxed is +8, or ~18 on average. Against a CR 5 target with 0 ranks in Sense Motive, the typical check is +6 for bab. This is pretty even given the Rogue has maxed skill ranks in Bluff. However, against a Djinni, with +12 for Sense Motive and +7 for base attack bonus, the check is +19 or ~29 on average. My assertion is that this is pretty unbalanced, considering both parties maxed out ranks in the opposing skills.
Given that assertion... What to do about it...
If you allow the feinter to add base attack bonus, this balances against Sense Motive, but then makes the feinter a shoe in versus target's without skill ranks in Sense Motive (which is most targets).
I'm tempted to house rule the target's check is the better of bab or Sense Motive, but still seems a bit off.
Demoralizing an Opponent is a bit awkward, but doesn't suffer this same swinginess because the target gets no ranks to add.
Per KRyan's comment, I have also considered the house rule of making a Feint in combat a move action, which goes to a swift action with the Improved Feint feat, and affects the all attacks against the target by the feinter until the feinter's next turn.
This would make feint more viable (and worth feat investment). I don't think it has a great impact on balance, since the alternative is flanking. It just opens up different strategies, and with a feat investment, counters the free flanking with cooperation. I'd appreciate arguments in support or against this.
Now that I think about it more, giving the feinter the bab bonus but NOT changing the action investment is fairly reasonable.. it's a shoe in most of the time, but costs an action and only affects one attack. You could add a "Greater Feint" which makes it a swift action that affects all targets for one round, and makes it a good, but feat costly strategy.