Since there's obviously something wrong if I found that rapiers/scimitars are better than heavy picks for a FB.
Rapier/scimitar (keen): 1d6 base damage, 15-18/*2, 19-20/*5
Heavy Pick (keen): 1d6 base damage, 19-20/*7
So, because the base damage on both is the same, we can just calculate which weapon a +1 damage modifier favours.
Rapier/scimitar (per 20 hits):
1: 0
2-14: 13
15-18: 4*2 = 8
19-20: 5*2 = 10
Total per 20 hits: 31
Heavy Pick (per 20 hits):
1: 0
2-18: 17
19-20: 7*2 = 14
Total per 20 hits: 31
So a if +1 increment to damage is the same for both, then a +n increment to damage is also the same for both.
(Even against fortification, they're the same; remember not to simply modify the damage they get on crits by the fortification, since they always get at least +1 damage. For example, a heavy pick crit under 50% fortification would not deal 50%(7) = 3.5. It would deal 1 + 50%(6) = 4, since the 1 damage is still applied on a negated crit).
(In fact, a quicker way to do this would to just measure what I'll call "crit addition" - how much more damage a weapon deals over just a flat +1 each hit for ever +1 damage added (i.e. no crits). To calculate it, just do threat*(multiplier-1), because we can assume a weapon which just hits 1 each hit has a multiplier of *1. So a rapier/scimitar would have 12, and so would a heavy pick. This makes it clear that difference fortifications has no affect on the difference in damage)
Furthermore, bursts affect both equally. A heavy pick's burst is 3d10, while a rapier's/scimitar's is only 1d10. However, a rapier crits 6 times, so cumulatively it is 6d10, while a heavy pick gets only two, for 6d10 as well.
(Remember that bursts apply normally, even on fortification)
Up until now, it looks like both weapons are in fact identical. But, there's one catch: seeker gives a minor boost to rapiers/scimitars.
Over 20 hits, +1 seeker damage gives 4*2*1 + 2*5*1 = 18 more damage to a rapier, while only 2*7 = 14 more damage to a heavy pick.
Granted, it's only 0.2 damage per swing more per seeker damage (so that's 1.2 if using a bloodstone), so it's a minor difference. But still, conventional wisdom says I'm wrong, so I had to have made a mistake somewhere.