> If we are in search of truth and want to build a consistent model of the world then we just can't accept this kind of reasoning.
Yes, you can. It's called generalizing. Remember, we're not doing math here, this is ethics.
Think about all the different situations where being treated equally to someone else is a good thing. Now you generalize that into "everyone is created equal", and that turns into a justification for all these situations.
If you want to criticise the generalization into A, there is a way:
Construct some C so that
1) A => C
2) C is not beneficial
Alternatively, you can attack the implication "A => B" or you can question whether or not B is beneficial.
The valid generalization would be "everyone should be treated equally" not "everyone is created equal" which is a totally different thing.
EDIT: I think what you're doing here is criticising A by constructing A' and arguing that it is better (by having the same consequences while being less general). That is a good way of criticising the generalization as well.
> Yes, you can. It's called generalizing. Remember, we're not doing math here, this is ethics.
To strive for a consistent set of beliefs grounded in reality is more akin to most branches of Philosophy than Mathematics as far as I am concerned, and seems particularly concerning to Ethics (e.g. you can achieve a set of beliefs of what you can consider 'good' or 'evil' ignoring all perception of reality but would the result be desirable?).