Here is the exact calculation with explanaitions.
On 6 decks (312 cards) we have the probability of a 3:2-payed blackjack:
(2 * 24/312 * 96/311) * (1 - 2 * 23/310 * 95/309) = 4.5323%.
The first factor is getting dealt a Blackjack (roughly 8/169). The second factor is the probability the doesn't get a blackjack (roughly 1 - 8/169) conditioned that you have a blackjack (1 ace and 1 ten missing).
As all suits are equal regarging a normal blackjack, there are 4*4 = 16 different suit combinations, 12 are unsuited, and 4 are suited, obviously 1 of them is the hearts BJ. We only get the bonus on hearts, so we divide the probability by 16, to get the probability of a hearts-suited BJ (while dealer has no blackjack):
4.5323% / 16 = 0.2833%
The bonus pays 2:1 instead of 3:2, which is half (50%) the initial wager.
Hence the player advantage (per initial wager) is
0.2833% * 50% = 0.1416%.
If the bonus is also paid when the BJ pushes (dealer also has blackjack), then we need to add things up:
Probability of a BJ push: (2 * 24/312 * 96/311) * (2 * 23/310 * 95/309) = 0.2166%
A BJ push while player has a hearts-suited:
0.2166% / 16 = 0.0135%
Normal pay is 1:1, if bonus pays 2:1, this is a full wager (100%), hence
0.0135% * 100% = 0.0135%
We add both up (since they are mutual excluded):
0.1416% + 0.0135% = 0.1552%
Of course a side count would help. If you can only keep 1 side count, I guess the Aces of Hearts are best indicators. More accurate would be two side counts with Aces and Tens of Hearts of course. One should do simulations regarding indexes and probabilities.