Would killing Putin be good or bad? Deontology and Teleology both have their answer set in stone, and are pure opposites. For them, the answer is 100% objective in the end. If you follow pure Deontology, killing people is objectivly bad, even if it's Putin, and if you are a Teleologist, you will say that it's objectivly good, because the result is good.
But of course, in the ethical sense, we only get more questions: How bad is "killing is bad" and how good is "Russia becomes better"? What if we are mistaken, and the succesor of Putin is even worse? How do we measure the risks in the result? What about the risks in the action?
Ethics tries to deal with all of these questions, and gives the people a s***ton of rules and tools and concepts to try to define some stuff. Some of these rules are even objectivly true. Like, actions which can be reversed are always better than actions which can't be undone if they turned out bad. That's mostly the case for binary checkboxes. "Can be undone" is simple and doesn't have to be quantified at all.
The problem is that "how much better is "can be undone"" is subjective again, and you can't call every action bad, just because it's not possible to undo them.
Now, the concept of Utilitarianism is simple. Good is, what creates the greatest happiness for most people, and the smallest unhappiness for the fewest people.
That's the concept and... that's it. Everything that follows it is again 100% subjective.
It doesn't define happiness, unhappiness, benefit or whatever at all, we have to define this on our own.
What is better? A little happiness for many or big happiness for a few?
Is big happiness for many, with some unhappiness for the few, better than a little happiness for many, without anyone becoming unhappy?
Utilitarianism has no answer to that, because you can't quantify or measure happiness at all. Good job, we are back to subjectivity.
In Statistics, the decision to reject the null or fail to reject the null hypothesis is 100% subjective, and as you laid here and in the rest of your post, the moral decisions we make is 100% subjective. However in Statistics there is also an objectively correct answer. The null hypothesis is either correct or wrong. In A LOT of scenarios it is physically impossible for you to know the correct value of the parameter. The error in the data, the random affect of probability, the arbitrary significance value cutoff (typically 5%) all make the reject the null hypothesis or not a subjective decision. However none of that subjectivity erases the objectivity of the parameter's value. The parameter has an objectively true value but because of error and uncertainty we are making subjective guesses at its value. Same with morality, if there is a correct answer on if we should kill Putin or not, then there's objectivity, and our failure to adequately define it and our failure to know the correct answer does not change that objectivity just as our failure to adequately measure the parameter does not change its objectively true yet unknowable value.
---
"Something's wrong! Murder isn't working and that's all we're good at." ~Futurama