Here are a few more thoughts on how best to interpret 'equal concern'. I think this might best be understood as a matter of procedural fairness. We establish a fair process, and accept whatever results it yields. ('Substantive' fairness, by contrast, requires a particular result.)
So what would a fair procedure be? It needs to take everybody's interests into account, and treat them equally, i.e. without favouritism. The Rawlsian 'Veil of Ignorance' springs to mind. Imagine that everyone affected is put behind the VoI, where they no longer know any of their own personal details (e.g. wealth, age, sex, education or talents). They then get to deliberate together and choose the basic structure of society -- a decision that aims at self-interest, but will be untainted by personal bias, since for all they know, they might end up as anybody in that society.
Now, suppose they had two choices:
1) Everyone has mediocre well-being
2) A small proportion of people are moderately poorly off, but everyone else flourishes.
Surely the rational thing to do is choose #2? But if this is so, and if the VoI - as described - really is a fair procedure, then this establishes that utilitarianism, not egalitarianism, is the theory which treats all people with equal concern.
Note that Rawls only manages to get his 'maximin' principle out of the VoI procedure by insisting that the relative sizes of various social groups are hidden behind the veil. In other words, the choices would be:
1a) Everyone has mediocre well-being
2a) There is a group of moderately poorly off people, and a group of flourishing people.
Here it may indeed be rational to choose #1a, simply because one cannot accurately assess the risks involved in #2a. But what justification is there for hiding this information? I think the process is fairer when the relative probabilities are made known. If we do not know how many people are in each group, how can we weigh their interests fairly and appropriately?
If done properly, it seems to me that the VoI becomes equivalent to the 'ideal observer' theory. That is, we imagine an omniscient, benevolent, impartial spectator, and ask them which option would be best. It seems likely that such a procedure would favour utilitarian over egalitarian outcomes. So that's what equal concern requires.