2nd edition
The Vf of these LEDs is only 1.95 nom@5mA.
ESR= ΔV/ΔI= 20Ω below 5mA and 10Ω above , not 100Ω
Using the Vt or threshold voltage at 1mA V=1.86, I have estimated the ESR from above. To drop from 3.3V to 1.95, current limiting resistors may be selected based on ESR of 20Ω for 5mA.
If you put 6 LED in parallel , there is a low probability of thermal runaway. But above 5mA the knee of the curve varies for different batches and hence there is a wide variation in Vf min-max @ 10mA.
The safer design isolates diodes with a small R to normalize the variation in ESR. then one R may be used to change the current in all of them (7R's), or use 6 fixed R's to limit 5mA in each.
To drop 30mA from 3.3 to 1.95 requires 45Ω.
You may test 6 in parallel with one 45Ω resistor will see even brightness but I cannot guarantee the margin to thermal runaway* without more data from Panasonic.
* Thermal runaway is a mechanism when power semiconductors with NTC voltage drop are put in parallel. The device with the lowest ESR draws more current. If the power dissipation is enough to reduce ESR or Vf then current becomes hogged by device. Generally it is not a problem in low power devices and perhaps not here either, but without more calculations with supplier tolerances,
In 5mm 30 deg 590nmD LEDs, I get parts with over 15,000 mcd @ 20mA in bulk orders, whereas this design is only 6x 30 mcd @ 30mA.
Variations of this suggestion include adding 10Ω in series to each LED with 43Ω to feed this parallel array. The 10Ω raises the voltage from 1.95 to 2.00 and reduces the variation in each string. The 43Ω @30mA=1.3 accounts for the drop from 3.3 to 2.0