Replacement theory is like any other theory - it makes
predictions, based on (a) some sort of model put forth (b) observations are made to test the model.
So for example consider the model "Fs = kX" for a spring. To test that model you apply forces to a spring, measure the extension, and see do they conform to your model.
In a similar manner the typical model for replacement theory is a continuous linear increase of various ethnicities and cultures of non-white persons introduced into a population, up until the point that the white population becomes less than 50% of the total population, at which point a postulated apocalyptic situation occurs.
You take this postulated linear growth and you are able to predict an approximate year the great event is going to occur (further you are predicting that the event that will occur shall be apocalyptic, based on another mental model).
Consider that this theory differs from Hooke's law, in that we can't test the predictions of the model until they happen in the future, as alleged.
All the theory has to go on is selective snapshots of the growth curve postulated by the model, alleged to be linear (but that is only because you are only looking at a small part of the total growth curve of course).
So first we're lacking adequate observations to remove uncertainty about the set of possibilities. Second, as I have said, we never see linear growth curves in nature, they are S-shaped curves.
And you could show up the theory as farcical in more ways if you had the inclination.
Granted, the term "conspiracy theory" usually refers to the introduction of scheming Jews driving some hypothetico-deductive scenario, and in the above, it is true that many proponents of replacement theory often do that.
But if they don't, what do we refer to the theory as? A brain-fart? Or something else?