No, not at all. You can't feel some way that's "incorrect" (or correct). Even if the way that you feel is somehow contradictory (relative to other ways that you feel, for example). However you feel is how you feel at a given moment, and there's not a correct or incorrect way to feel.
Even a moral relativist should answer the question: How is the moral culpuability different in the three scenarios I suggest?
Yeah, I can answer that, although I'm not sure what it's going to matter.
In (1) Actions do not have to be direct. We can be talking about a causal chain of events that rather indirectly lead to nonconsensual harm, whether we're talking about something intentional or accidental/negligent. This is a case of an indirect action. In this case we're talking about something intentional, so it would be punished more severely than if it were negligence.
Re (2) The child is responsible for the bombing, but the bombing isn't an intentional act on the child's part, and the child probably wouldn't face repercussions--there would likely be no reason to expect a child to be skeptical of pushing a plunger. There's a matter of contractual fraud in this scenario, because the results of pushing the plunger are not being disclosed, even though they're known. Cases where contractual fraud leads to something like a death would be punished more severely if I were king than cases where contractual fraud leads to, say, losing $10 or whatever. The punishment for contractual fraud would be proportionate to the upshots of the contractual fraud.
In (3), assuming the accomplice knows the results of pushing the plunger (otherwise this is the same as (2)), the accomplice is responsible for the murder and the person who built the bomb is guilty of nothing more than building a device that might be prohibited (it depends exactly on the nature of the bomb, whether the builder is authorized to build such a device, etc.). In this case, it's the accomplice's decision to commit the murder that matters on that end.
In (2) and (3) we're not talking about a causal chain of events as we're talking about in (1), because we're positing people making decisions to do something as an intermediate step. That breaks the causal chain, as someone could decide to not push the plunger. And then it's just an issue of whether they know what the result of their decisions will be.