For a long time, socialists have been saying that 'real socialism has never been tried out before; when real socialism is fully implemented, it will produce amazing results.' This is a line of reasoning that can be observed when people talk about Venezuela, which suddenly was not a communist country at all (see Kristian Niemietz series about this). This presents us with an interesting epistemic problem: in what sense are economic theories testable? Is it necessary to create, for example, a pure communist system before we can criticise communist economics for being harmful? Similarly, do we need to have a fully capitalist system before we can know that capitalism works? Here I will present two answers to those questions.
First of all, economics is driven by the search for good explanations. When an economic theory is an inadequate explanation, we can discard it (see Brett Hall's description of a good explanation). For example, there are sound theoretical arguments against communism, on the sole basis of which we can already reject Marxism.
However, there is also a second answer. To the extent that economics is testable, it is testable through the incremental application of its theories. Whenever free-marketeers claim that capitalism is better than communism, what they mean is that the world would be a better place with more free markets. The argument that they are making is that less regulation creates wealth, and more regulation decreases wealth. These are testable predictions.
Similarly, Marxism, socialism, and communism are testable to the extent that we can make a move towards them: for example, even when these ideologies are only partially implemented, the underlying claim that more state control is better *is* testable. In this sense, we resolve the unfalsifiability of economic theories.
It should be noted that this necessarily true for a good economic explanation. An economic theory that only works in the abstract is not implementable, and we should not try to implement it for that reason. Now, I am sure that people will argue over whether communism or capitalism is implementable in the above-described way. Nonetheless, even in the case that neither communism nor capitalism is achievable, any sound economic theory has to be incrementally implementable. For otherwise, it would not be implementable at all, and it would not be a good economic theory. In this way, any sound economic explanation will be experimentally distinguishable from a bad one.
There is an addendum: we should note that any experimental test is itself filled with explanations, which in turn need to be good. It could, for example, be the case that government control seems to increase wealth creation, when in fact we are observing a market effect. The previous instance examplifies why falsifiability is not the main aim of science.
In short, we should not have to worry about whether or not 'true' communism or capitalism have or will ever be tried out. Any good theory will reveal its worth when fallible humans try it out, with all the errors and parochialism that comes along with that.