BLM

[Jelulish]

Yeah I tested it extensively....and in short you are absolutely wrong.

Here is how I tested it:

1- Went to AK with equally geared BLM.
2- We both agreed on using the same exact rotation until the first boss (bliz 3 > Thunder 3 > Fire 3 > Fire 1 spam and use Thundercloud and Fire starter).
3- We both did 111k.
4- After killing the first boss I switched to your method until the second boss, while he used the exact same normal rotation.

The end result was ugly I did 16k less damage than him. I have also tested it on Turn 1 and same result I preformed poorly when I'm used to be the top DD in my Coil group.

Math is not absolutely wrong, it's a 7.56% dif any way you wish to put it. With the fire 1 rotation you can get higher or lower depending on how lucky you get with your procs but on average its a 7.56% dif. So, fire 3 is slightly less efficient I will concede to that.
(((150*10)+((220*10)*0.4))-(220*10))/(((150*10)+((220*10)*0.4)))*100

[Thane]

Math is not absolutely wrong, it's a 7.56% dif any way you wish to put it. With the fire 1 rotation you can get higher or lower depending on how lucky you get with your procs but on average its a 7.56% dif. So, fire 3 is slightly less efficient I will concede to that.
(((150*10)+((220*10)*0.4))-(220*10))/(((150*10)+((220*10)*0.4)))*100

It should be noted that Fire I is far more consistent than Fire III (shorter cast time, less mana), and it provides means of doing large amounts of damage on the move via Firestarter procs. I can see an argument for using Fire III over Fire I if you absolutely never have to move during an encounter, but we both know that condition is almost never met. I'm not sure if you're basing your conclusions solely vs Firestarter procs, or practicality. In a real "world" situation, Fire I is the better choice.