In principle, I agree with Josh Fershee's analysis of opting out of the fiduciary duty of loyalty in the LLC setting:
At formation, then, those creating an LLC would be allowed to do whatever they want to set their fiduciary duties, up to and including eliminating the consequences for breaches of the duty of loyalty. This is part of the bargain, and any member who does not agree to the terms need not become a member. Any member who joins the LLC after formation is then on notice (perhaps even with an affirmative disclosure requirement) that the duty of loyalty has been modified or eliminated. This is not especially concerning to me.
What would concern me more is a change in the duty of loyalty after one becomes a member. That is, if the majority of LLC members could later change the loyalty provision, then that seems problematic to me, as fiduciary duties are not just to protect the majority. As such, it seems to me more proper that changes to the duty of loyalty, when a member does not have any say in that change, is what should be restricted. Like in changing a partnership agreement, if everyone agrees, then there is not a problem. And if you accept the provision when you join, it is not a problem. But you shouldn't have a fiduciary duty removed or modified after the fact without your consent.
To the extent that we're talking about LLCs of the mom and pop variety, I agree. But what if the LLC has 50 members? Or an LLC that's publicly held with potentially thousands of investors? In such firms, unanimity rules create serious holdout problems. (Of course, you can get holdout issues in the smallest LLCs too, but as firms get larger the odds of a holdout go up and the social norms that prevent holding out weaken.)
Lucian Bebchuk nicely set up the issues in The Debate on Contractual Freedom in Corporate Law, 89 Colum. L. Rev. 1395, 1400-01 (1989):
Unlike initial charters, charter amendments cannot be viewed as contracts; consequently, one cannot rely on the presence of a contracting mechanism as the basis for upholding opt-out charter amendments. Furthermore, the amendment process is quite imperfect and cannot be relied on to preclude value-decreasing amendments. Although an amendment requires majority approval by the shareholders, voting shareholders do not have sufficient incentive to become informed. And although the amendment must be proposed by the board, the directors' decision might be shaped not only by the desire to maximize corporate value but also by the different interests of officers and dominant shareholders.
Rational and informed shareholders forming a corporation would recognize the desirability of an amendment procedure that permits charter changes without unanimous shareholder consent. But they also would recognize that allowing any given opt-out freedom might produce value-decreasing amendments and thus involve an expected cost. The expected cost of a given opting-out freedom, as well as its expected benefit, would vary substantially depending on the type of issue involved and the circumstances under which the opting out would be done. An analysis of these expected costs and benefits suggests that the optimal arrangement-the one that rational and informed shareholders would wish to govern their future relationship-is one that, while allowing much midstream opting out, also places significant limits on it.
The optimal arrangement thus involves an element of precommitment not to adopt in midstream, at least not in certain circumstances, opt-out provisions with respect to certain issues. This optimal arrangement is the one that the law should provide, at least in the absence of an explicit provision to the contrary in the initial charter. Even strong believers in free markets should accept this optimal arrangement as the default arrangement in the very common case in which such an explicit contrary provision is absent. Therefore, even supporters of a complete freedom to opt out in the initial charter, I suggest, should recognize that mandatory rules have a desirable role to play in midstream.
But what mandatory rule? How about the one formerly adopted by the MBCA with respect to midstream amendments to the articles of incorporation? I.e., a modified supermajority vote (perhaps a majority of the total number of members--not just those present and voting at the meeting--or, if they have unequal voting rights, a vote of the holders of a majority of the outstanding voting power) coupled with a buyout for dissenting members?