Basically all of astrophysics "suffers" from the falsifiability problem – yes, careful observation can yield results that parallel falsifiability but you can't really create a circumstance where all the variables are accounted for in order to establish falsifiability in astrophysics; it's not like you can blow up the moon just to prove that it affects the tides on earth or to create a second universe to tinker around in, you know?

Sciences like evolutionary biology and geology do not confine themselves strictly to the scientific method as it's not possible to apply this to all areas of these sciences by their very nature. (Sure, you can do a lot with fruit flies or microbes but it's not possible to apply the scientific method to extinct animals and to our ancestors from thousands or millions of years ago.)

There are plenty of examples of scientific discoveries that occur accidentally and I think it's fair to argue that Darwin's theory of evolution falls outside of the strict definition of scientific method as he had plenty of observations but he didn't observe these changes directly but merely the end result of these changes, so it's something that sits outside/alongside the scientific method rather than fitting neatly within it.

I know that what I'm saying could be interpreted as a very naive dismissal or rejection of science because this comment gives the appearance of me undermining or denying the validity of science but that's really not the case at all even though you can hear these arguments from, say, more sophisticated creationist charlatans who point out that there are parts of evolutionary biology that are not falsifiable and then go on to argue that therefore it's not a science but a myth and blah blah blah. My argument is the polar opposite though. Evolutionary biology has a staggering degree of descriptive and predictive power despite having one foot outside the strictest definition of science but that means that either the way that we define science is faulty or there are factors that exist outside of that definition which are critical to the existence of science, not that there's something wrong with evolutionary biology or that we should discredit it. There's the famous case of biologist Richard Alexander predicting the existence of naked mole rats with remarkable precision based purely on extrapolation and hypothesising, for example. There's no denying that this sort of extrapolation is one of the best ways to vindicate evolutionary biology, outside of what we are capable of directly observing with short-lived organisms.

So if we were to be strict about how we apply labels, what Richard Alexander did was undeniably scientific however it wasn't itself science. And I think that Marxism is very much like this. Moreover, Marxism is not doctrinaire. People can be dogmatic Marxists or class-reductionist, sure, but there are plenty of examples of people being dogmatic about a science as well and that doesn't disprove the science itself it simply shows that people are fallible. Again, this is another thing that anti-science people point to and they try to frame a scientific discipline as being static and fully realised so then if someone is being dogmatic about science they use that to claim that therefore something is fundamentally wrong with science itself. Meanwhile, scientific debates are an everyday occurrence, they are often heated, and it's not uncommon for there to be factions that battle out different hypotheses amongst themselves because a major aspect of science is its ability to incorporate new information and to adjust to changes in how things are understood and framed, and these debates are a major part of this process. In a similar vein, Marx would have been looking to Britain, Germany, and the US for revolution to occur based on his analysis. Then we see that the antagonisms between classes in these most industrialised countries being transposed onto other countries, essentially overcoming the factors that hemmed in capitalism within that nation's borders without actually resolving them because they were effectively outsourced to other countries. And that's where Marxists observe this phenomenon and they develop a theory of imperialism that accounts for these new developments and which points to the existence of different fault lines within the capitalist world order that are vindicated by later revolutions as it proves that revolutions are more likely to occur in the countries where capitalism is weakest, in the margins of the hegemony of global capitalism, rather than simply happening in the most developed countries.

So Marxism incorporates this and develops onwards, and to me that is sufficient to call Marxism scientific because it makes observations about history and the present, it collects data about what is going on, it creates a theory based on those observations, this is extrapolated out into hypotheses, we observe whether those hypotheses are in line with what society produces, we debate the results and the causes, and where there are outlier events like revolutions occurring in countries like Russia and China and Vietnam we go back over our observations and hypotheses to adapt to this new information. Or at least some of us do anyway lol. And the same can be said of tactics and strategies for the ol' educate/agitate/organise - most Marxists aren't following the Bolshevik revolution as if it were a step-by-step process but rather it's about identifying the similarities and the differences then applying the lessons from historical revolutions to bring about the next one however we can.

That's sufficient to call it a scientific approach in my mind. If you want to call it a science and to take a broader definition of what science is then I'm not gonna object to that. As far as I'm concerned the debate over whether Marxism is itself a science is a lot like a border dispute - people are arguing over what lies within a territory and what falls outside of it. This can be a productive discussion but I'm much more interested in examining the topography of the map than discussing the borders because that's what yields the most important information and it has the most direct relevance to application imo.