You can steal a bank and perhaps you can help more people than the bank that initially had the money, that is clear. Breaking rules can often be positive. But that's not the point.
Rules exist in order to allow healthy interactions (for a theory of interactions, check the links on my profile). Usually, a healthy interaction is the one that produces constructive results for both interactors (e.g. you solve a problem with a creative agreement, not by killing the other). Ideally, a rule that generates a common positive interactional result should never be broken, because it will always be better than breaking it.
The problem about assessing interactions is its systemic nature. The results of a rule are propagated throughout thousands of subsequent interactions. When society creates rules, we try to estimate all their impact. That is the reason it is very difficult to define "good" rules. Society just do its best in establishing rules. But the effect of rules is usually only observable with time.
In consequence, some rules can be "negative". In addition, most rules can cause advantages for some and negative for others. But none of this two arguments is a valid justification to break the law. Then, even if we die, it is correct to respect law, because it is us who define them. That does not mean at all not to be reactive against negative laws. If time shows that a law is negative, it is our duty to fix and change it. Of course, our sociopolitical dynamics force us sometimes to get divided in order for bad rules to be eliminated. Divisions are always destructive, cause negative interactional results, so that is the way that normative issues are naturally forced to be solved.
Breaking rules is incoherent, and not socially desired. Laws being broken is a signal to pay attention, and enforce or change them.