Ifrit-egi is bugged.

[Taranok]

Just tested it myself, i444 (bis, half materia). Without crit/dh, Ifrit had an auto attack damage of 1800, garuda 1530, and titan 1400. Numbers are approximate and rounded to nearest hundred, bumped garuda's because it was noticeably above 1500 every time.

There is definitely something wrong here, even if you assume pet scaling of 80% SMN stats, it should still be consistent within itself of Ifrit > Titan > Garuda at 2x, 1.5x, and 1x garuda's resting potency, which it clearly is not. Ifrit is definitely stronger, but not twice as strong, so something is screwy here.

As a note, Burning Strike, Wind Blade, and Rock Buster are the pet's "auto attackls." They do not hit outside of these named abilities automatically.

For reference, Ifrit is hitting for 20% more than Garuda if I assume garuda hit at 1500 at rest, and Titan for just under 7% less than Garuda. Remember, Ifrit should be twice as strong as Garuda, and Titan 50% stronger. Crimson Cyclone appears to have hit for 2300-2400 potency, flaming crush for the same 2300-2400, 30% stronger than 80 potency when it should be 25% stronger, but that's within margin of error.

Slipstream, at 20 potency, hit for 800, putting it at 48% less damage than the 40 potency auto attack (margin of error). Meanwhile, Aerial Slash, 60 potency, was hitting for 2400, 56% more (margin of error) making it consistent with the 40 potency auto, but also meaning 60 potency on Garuda is equivalent to 100 potency on Ifrit, because reasons. This is also true with the 20 potency Slipstream, it's 1/3rd of the 60 potency Aerial Slash.

Meanwhile, Titan's Mountain Buster at 70 potency comes in swinging for 1600, an 11% increase over its resting state of 1400. For reference it should be 16%, so again it's within margin of error for such low sample sizes because laziness.

So, all 3 pets seem to be consistent within themselves, but the relationship between all 3 of them is completely broken.

As a control, Ruin 3, at a default of 100 potency, swings in for 4500-4800, meaning that Ifrit is at about 39 potency, whereas Garuda is at 33 potency, and Titan at 30 potency for its main attack, and everything else is relative to that. Again, low sample sizes. But the most obvious observation is that Ifrit is less than half as strong as it's supposed to be, Garuda at 75% as strong as it's "supposed" to be, and Titan at half as well. So Garuda is definitely overscaled, but the pet penalty is ridiculous to say the least. This is insultingly low and I'd seriously consider deleting the egis because these numbers are that bad.