I'd need someone to spiritbond 3 pieces of gear with identical iLevel (easiest is probably i55 gear bonded while killing open-world shit) while wearing as much +spiritbond as possible. Assuming you had all +spiritbond slots occupied with +spiritbond gear, I would suggest bonding i55 hands, i55 legs, and i55 feet. Make sure that the i55 hands are unmelded and NQ, the i55 legs are HQ and single-melded, and the i55 feet are HQ and triple-melded. The only data points I need are
1) The assurance that these pieces of gear were all bonded simultaneously, with no variations in party size or amount of +spiritbond.
2) The amount of +spiritbond used (the more, the better)
3) The amount of bond% on each piece of gear when the feet (or whatever piece is HQ-triple-melded) reach exactly 100% bond
From there, I should be able to figure out the potency of the +spiritbond stat.
"How could that possibly be enough data to reach a conclusion?"
I know that an HQ-triple-melded piece of gear bonds at exactly twice the speed of something that is NQ-unmleded, and that an HQ-single-melded piece of gear bonds at 150% the speed of something that is NQ-unmelded (because each HQ/materia is worth +25% bond rate). Using this data provided, it could show slight differences in these percentages that would reveal how potent a point of +spiritbond is.
With +0 spiritbond, once the HQ-triple-melded piece reaches 100% bond rate, you should see...
HQ-triple-melded: 100% bond (bond rate of 200%)
HQ-single-melded: ~75%* bond (bond rate of 150%) [(100/200) * 1.50 = .75]
NQ-unmelded: ~50%* bond (bond rate of 100%) [(100/200) * 1.00 = .50]
If +14 spiritbond is as simple as +14% spiritbond, then you might see...
HQ-triple-melded: 100% bond (bond rate of 214%)
HQ-single-melded: ~76%* bond (bond rate of 164%) [(100/214) * 1.64 = .766355...]
NQ-unmelded: ~53%* bond (bond rate of 114%) [(100/214) * 1.14 = .532710...]
However, if the +spiritbond is taken into the equation AFTER all the HQ/materia bonuses have modified the bond rate, then your numbers would look exactly like example 1 (100/75/50 bond rates), and we would need to proceed to the cringe-inducing alternate method (see below).
*Give-or-take 1% due to potential rounding errors. Hopefully rounding errors don't bump that % up any more than 1, but I don't actually know.
Note: If you can't muster at least +10 spiritbond, don't bother trying. These numbers are so close to one another that it's difficult to account for rounding error if the +spiritbond number isn't as large as possible.
Alternate method, for those with a lot of patience:
Follow the above instructions, but instead, keep track of exactly how many monsters is takes to make each piece reach 100%, with some additional rules.
• Count EXACTLY how many monsters it takes for each piece to reach 100%.
• Make sure that you only ever kill one type of monster (I suggest the hippogryphs in mor dhona).
• DO NOT do this in a party, since I don't know exactly how the party size effects spiritbond (this includes your chocobo).
• DO NOT kill monsters that link (I don't know if this affects spiritbond or not).
You're not done yet:
• Now do the exact same test with the exact same conditions, but with +0 spiritbond, and compare the amount of kills it took to reach 100% on everything.
I tried to do this once, but I kept losing count because it was boring. For sanity's sake, I do not suggest it.