Tested with a friend tonight. I wasn't able to completely flush out the mechanics of this yet (you'll see why). For the sake of 99% of the readers that just want the conclusions, I'm just going to post results first, then a separate section with the math / methodology for the few that don't mind checking for me.
Rule of thumb - you can get by just reading the bold/red text.
For reference:
Yonin = + enmity
Innin = - enmity
Both JAs share the same timer. They wear off after 5 minutes, which is also the recast timer of their use.
****************************
Results
CE and VE values generated by the act of using the JA itself:
Yonin = 1 CE | 600 VE
Innin = 1 CE | 80 VE
* The amount of +/- enmity granted from Yonin/Innin is additive with your gear and merits. For example, if Yonin gives you +20 enmity, you have +4 from merits and +10 from gear, your total is basically +34.
* The amount of +/- enmity granted from Yonin/Innin changes over time.
* I estimate (I stress, estimate!) that Yonin starts at +30 and eventually wears at +10 after 5 minutes. Innin around -20 at start. Currently not enough good data to suggest an amount near the end.
At this point, you can stop reading the parts that are not bolded if you just want the conclusions/
****************************
Methodology
RDM/BRD (0 enmity merit) and NIN/RDM (+2 merit)
NIN opens with Slow pull (205 CE)
RDM cures NIN 5 times, then songs both players 5 times (205 CE) - even CE.
RDM cures himself twice (+2 over NIN)
NIN uses JA, looking at recast timer for number of seconds to switch back.
NIN uses cure once to verify the 1 CE from the JA (successful)
Recast timer showed roughly 9-10 seconds to switch target back. I rounded this to an estimated 600 VE (60VE/sec*10sec).
***************
For the enmity change effect, we used the same mob, even out CE before beginning. I'll start with Yonin
(RDM 0 | NIN 0) : Start even CE (balanced to 0)
(RDM 1 | NIN 0) : RDM uses cure on himself once for 0 HP.
(RDM 1 | NIN 1) : NIN uses Yonin.
------- wait X amount of time before continuing, hate rebalanced -------
(RDM 0 | NIN X) : NIN uses Dispel.
At this point, RDM uses a combo of BRD songs and cures to regain hate. Here are the results for 3 trials at separate time. Timestamp was used to determine time between the usage of the NIN JA and the time of the Dispel.
Trial I:
NIN enmity = +2
Time from JA use to Dispel = 87 seconds
CE generated from Dispel = 403 CE
At this point, I did not know if the enmity enhancement of Yonin was added to gear/merits (like Dirge) or multiplied (like SCH equanimity). This test suggested to me it was additive based on this math:
403/320 = 1.2594 | 404/320 = 1.2635 (upper bound based on truncation rounding)
Any value that falls between 1.2594 and 1.2635 is valid. Obviously 1.26 is the most friendly number for addition (can only result in whole numbers). If we try the multiplication, then for some value X,
1.2594 < (100+X)/100*1.02 < 1.2635 , where X must be a whole number.
No whole number meets this criteria (closest are 23 and 24, but both fall out of range). Hence at this point, I feel it is safe to assume that the JAs are additive with merits and gear.
Thus, the result of this trial showed a +26 enmity enhancement in total. Subtracting out the +2 from merits, we reach the conclusion that...
After 87 seconds since Yonin use, the enmity enhancement is +24.
Trial II:
NIN enmity = +10
Time from JA use to Dispel = 39 seconds
CE generated from Dispel = 438 CE
I had my friend (NIN) now add some enmity gear to reach +10 from gear and merits. We then repeated using a different timing after JA usage (only 39 seconds this time). The 438 CE result results in this enmity enhancement range:
438/320 = 1.3688 | 439/320 = 1.3719
We find that 1.37 fits within the range, so at this point I assume the total enhancement is +37. Subtracting out the +10 from gear/merits...
After 39 seconds since Yonin use, the enmity enhancement is +27.
Trial III:
NIN enmity = +10
Time from JA use to Dispel = 275 seconds
CE generated from Dispel = 387 CE
Same thing but we waited 275 seconds (the effect wears at 300 seconds). We find that...
387/320 = 1.2094 | 388/320 = 1.2125
Again, a good number appears in range at 1.21, or +21 enmity total. Subtract out the +10 from gear and merits...
After 275 seconds since Yonin use, the enmity enhancement is +11.
***************
Test for Innin is almost the same so I'll skip the methodology and go to results.
Trial IV:
NIN enmity = +2 (Using Innin)
Time from JA use to Dispel = 43 seconds
CE generated from Dispel = 265 CE
We run the same sort of calculation to find the range...
265/320 = 0.8281 | 266/320 = 0.8313
A good number appears at 0.83 or -17 enmity. If we consider the fact we had +2 from merits,
After 43 seconds since Innin use, the enmity enhancement is -19.
Unfortunately I wasn't able to run more tests at this point. I'll try more in the near future.
****************************
Questions Still Unanswered
(1) How do these JAs work with the +/- enmity caps? This may be a hard question to answer since I don't believe it is possible for a NIN to reach +100 enmity even with Yonin (without resorting to Sentinel, which is unique anyways). The - enmity cap would similarly be a bit hard to get the gear for, in addition to probably not being all that relevant to the actual game setting.
(2) How exactly does the enhancement (+/-) depreciate over time? Every X number of seconds perhaps?
(3) What are the lower and upper bounds of the enhancements? Specifically right after you use the JA and right before it wears? Right now I estimate Yonin at +30 to +10 and Innin at -20 to UNKNOWN.
http://kanican.livejournal.com/19569.html
Enmity Table was updated to reflect the basic CE/VE of using the JAs, but I'm still holding off on their descriptions of the enmity enhancements until I flush it out completely and have it verified here.