I believe in this hand, you can look forward enough to get situations the ICM can evaluate. That is not always the case in a folding war.
If you fold the SB, you will have 100 on the BTN with the big blind all-in for 540. I think the ICM should be accurate enough at the end of the next hand, and so we can use the Nash calculator's evaluation of 21.15%.
If you call and lose, you get 0. Ties are negligible when you have a pocket pair like TT. (They would be much more common if you had a hand like A5.) If you call and double up to 800, then I think the ICM is again going to be reasonable at the end of the next hand, although I think it will tend to overstate your equity because you would post the big blind after that. Posting most of your stack as the big blind is not as big of a distortion when you are the shortest stack anyway. The Nash calculator's evaluation is that if you double up now,
BTN (Hero): 800
SB: 10660
BB: 540 (all-in)
you will have 30.36% equity.
A very interesting phenomenon that I think most people would not expect is that you are not supposed to play very tightly from the BTN in this second situation. I'll talk more about the multiway penalty later, but the shortest stack is heavily penalized if you play a 3-way pot all-in, and you should not be so averse to getting all-in as the second stack once the shortest stack is all-in by posting the big blind. So, the Nash calculator suggests that you should push 33% from the BTN, knowing that this will usually produce a 3-way all-in.
If you trust these numbers, then the equity you need against a random hand to call in the original hand is 21.15/30.36 = 69.66%. 99 has 72.06% and seems to be a call. 88 has 69.16% and appears to be a close fold.
Of course, you can modify these calculations in case you want to drill deeper into the second scenario, or to adjust the first one for the chance that the big stack will give the big blind a walk. I think the former will slightly increase the equity you need, while the latter will slightly decrease it.
An advantage of this method is that it is simple enough to apply it with other prize structures. Once you are down to 3-handed, a 50-30-20 structure is equivalent to a 75-25 structure. 70-30 and 65-35 structures are more common for 6-handed SNGs, and we can also look at a super satellite which pays 2 people, 50-50.
75-25: 13.54/21.72 = 62.34% 66+ A9s+ ATo+ KJs+
70-30: 16.07/24.6 = 65.32% 77+ AJs+ AKo
65:35: 18.61/27.5 = 67.67% 88+
60-40: 21.15/30.36 = 69.66 99+ (calculated above)
50-50: 26.22/35.73 = 73.38% TT+