This situation depends on the players and their calling ranges. In Nash (if everyone made the optimal decision) this would be a shove +0.06%
However this is most likely not the case. If they are tighter than Nash then it's even more +ev if they are calling looser than Nash then shoving is a mistake.
Most ICM calculators don't account for future hands and since you are in the BB next this needs to be added into the equation. I do this manually but there is a feature called FGS (Furture Game Simulations).
This big blind is 1/4 quater of your remaining stack so folding is suddenly not looking like the best option in fact it's probably going to lose you money in the long run folding here. In FGS KJo is now +0.82 using Nash ranges so now you have some breathing room in case they call wider than Nash.
Another unaccounted for variable is that if you get called by the big stack and win you have the chip lead which is worth a lot of value at this point. If you get called by the short stack you have him covered and you would take him out and have a stack big enough to contest 3 handed. Doubling through the BB would also be great leaving two short stacks to your left $$$$
I can only assume that in most situations the SB will try to fold you out which IMO would be a mistake but since this is most likely the case it also hurts you which is another reason to go ahead and jam BUT if the SB is a maniac then that changes everything. The more volatile a bubble is (Bad players who don't care about bubbling) the more incentive to pass up on marginal or negative spot you are forced to take.
Not sure if min raising would be reasonable since you lose some fold equity but you do represent a super strong holding but obv with the intention of raise folding to 2 or 3 all ins and obv snap off vs just one player since it's most likely he was never folding anyway. This could be optimal dpeending on the player types.