Hello mael and welcome,
In theory, hydrogen bubbles in a hybrid fuel grain would increase specific impulse. In practice it will not be possible to embed enough hydrogen to make a measurable increase. Consider, for example, a fuel grain 10 cm in diameter and 1 m long. Of course there must be some solid fuel between the bubbles, but even if it was all hydrogen (no solids at all) at 2 atmospheres pressure, that would correspond to approximately 0.7 grams of hydrogen. If the fuel grain was half hydrogen and half solid fuel by volume, you would have just 0.35 grams of hydrogen in the fuel grain.
Hydrogen bubbles would probably increase the regression rate, but it would be extremely difficult to get consistent results from one fuel grain to the next. The number and size of bubbles would have to be the same each time in order to get the same burn rate.
Sorry to burst your bubble (no pun intended
), but it is often the case that ideas in science sound great until the calculations are performed...
Best regards,
Terry