After showing the thunder the gem I posted, he went on to test different ratios of ORB gems in the "Tick" operation. They are as follows:
3 grade 8 ORB to 13 grade 8 OB (3/13) - same method with grade 5 gems
A=(T+T)+((T+g)+g)
B=((((((g+g)+g)+g)+g)+g)+g)
C=(g+g)
D=(((B+C)+C)+A)
4 grade 8 ORB to 12 grade 8 OB (4/12) - same method with grade 5 gems
A=(T+T)+((T+g)+g)
B=((((((T+g)+g)+g)+g)+g)+g)
C=(g+g)
D=(((B+C)+C)+A)
The control gem (2/14 ratio, method I used) at G11 (G14e) is -
7.26 chain hit length
Leeches 458.37 mana per hit
3/13 ratio:
7.98 chain hit length
Leeches 458.19 mana per hit
4/12 ratio:
8.29 chain hit length
Leeches 457.95 mana per hit
Bloodbound is the same across all gems, which is x1.84 (0.839 per hit level).
At grade 26 supergem -
2/14
15.19 chain hit length
Leeches 1,474,739.34 mana per hit
3/13
19.51 chain hit length
Leeches 1,474,739.29 mana per hit
Bloodbound increment is identical - +5.535 per hit level.
To prove that just adding a grade 1 red at the end of a 16 combine gem isn't the most optimal, here is a grade 41 supergem with a grade 1 red in the initial 32 combine and one where the grade 1 red is added at the end -
Red at start:
71.93 chain hit length
Leeches 7,665,361,431 mana per hit
Bloodbound - +36.54 per hit level
Red added at end:
37.92 chain hit length
Leeches 5,555,814,392 mana per hit
Bloodbound - +29.232 per hit level
Please test for all cases, but if others' tests turn out identical at 0 hits and 60 skill level for orange, red, black, and fusion, then it is as someone (ronin316?) said; order of operations matters.
I and thunder each share the credit for this proof equally.