I agree, but there is no point in defining real to encompass things we cannot know.
And so then we will add it to the set of known real things.
My definition is the real things are things that have some effect on other real things. I don't think using the word exist is useful.
I agree that it is reasonable to say "There are probably things we don't know about that interact with things we do know about." From that it follows that those things are real.
We need to differentiate between the definition of real and the sets of known real things and hypothesized real things.