Yes, exactly. It's like complaining that gravity isn't "real" because nobody's touched it. That may be so, but it's real enough in that we can observe its effects and use those observations to make accurate predictions about the behavior of objects in space. If meme theory can do the same for the behavior of ideas in social groups, it's real enough too.
I have not followed meme theory: Can it?
I think the answer to that would have been no as of 2005 (http://cfpm.org/jom-emit/2005/vol9/edmonds_b.html), but I haven't been following since then so there could have been advances made.
I have a hopefully close friend who is much smarter than me whose field of study is memetics, so I am certainly not going to chuck out the theory for personal reasons as well as a natural cautiousness around ideas I don't really understand.
I think it's important that they bring up Dennett specifically, who has strong opinions about consciousness and intentionality that appear on a surface reading like they might conflict with his take on memetic theory (http://www.consciousentities.com/alternatives.htm#memes).
Goodman juggles vastly different concepts of what a "meme" is without consistently distinguishing between them. He seems to dismiss memetics as an abstract model based on a lack of evidence for memes as physical entities, which, depending on how it's constructed, seems like either fringe absurdity (an idea particle) or a logical conclusion of materialism. And his question of whether an idea or its proponent "wins" is a poorly-defined non-sequitur.
The original question is bad, in my opinion. Assuming you acknowledge the existence of ideas, a better question would be "is meme theory a useful or accurate model of how ideas behave?"