Thanks for the questions. Certainly, crab predators would want to maximize their energy efficiency by getting the most food energy at little cost. The classic Elner and Hughes 1978 (and many studies afterwards) suggest that there is an optimal size of prey consumed based on the energy needed to capture/consume the prey versus the energy the predator gets from the prey. Often, many crab predators choose to consume smaller prey than they are capable because it costs less, particularly in terms of potential mechanical damage, and thus is more "energy efficient" (see papers by Juanes). So while the green crabs could potentially consume the largest mussels offered, in terms of their ability, the cost was such that they choose to consume smaller mussels.
In terms of the shift to even smaller mussels when the predators were combined - we propose that this is due to the green crab spending less time foraging and more time dealing with the Hemigrapsus crab. Although the Hemigrapsus crab was much smaller, it was also aggressive, and we have video recordings of the green crabs spending time keeping the Hemigrapsus crabs away. Larger mussels require both more handling time and effort to break, and the videos suggest that the green crabs did not have this time when the Hemigrapsus were present. We think this is what ultimately led to the downward shift in prey consumed. So in essence, while larger mussels are more difficult to crush in general, it was probably the decrease in potential handling time that set the upper limit of consumption in the combined experiments.