Can constraints against goodness maximizing actions be defended on grounds which suggests that such actions might illegitimately involve one using others without treating them as ends? What is the difference between treating someone as an end and merely using that person?
The answer to your first question is "yes". Consequentialism, from the perspective of deontological ethics, lacks justification. If you do accept the idea that a person is and always must be treated not "merely as a means to an end, but always at the same time as an end" (Kant, Groundwork of the Metaphysic of Morals), then you can't defend a moral concept of maximizing happiness (or the like), because this is a consequence of actions that might come into conflict with your premise. Easy example: ship sinks, there aren't enough lifeboats, do we risk our boat sinks, or do we drop somebody to drown? If one dies, the other live, happiness is maximized, but dropping a person to their freezing death conflicts with our moral premise that a person must be treated as an end. Consequentialism and Kant don't go along.
To understand what the Humanity Formula of the Categorical Imperative means I recommend reading point 6. of this article.