If you change the definition from "find" to "find and prove" it holds quite nicely (but doesn't sound as good). This works on the theory that there are no applicable, proven, scalable, repeatable business models. (Due to the presence of established competition)
Models proven to be un-scalable and or un-repeatable is trivially possible.
Models which are scalable and repeatable but un-proven can exist, the company will be executing the business model and in that they will be demonstrating the viability of it assuming the company will lives long enough that this is possible. (Good ideas can still fail if they hit an early setback and don't recover)
If the model is proven to be scalable and repeatable then there has to be someone who proved it to be so (And given the quality of the business model and its lack of glaring flaws they will still be around). With this in mind the original proving company is first to market and likely has a huge advantage, so unless the original product is markedly worse than the new one the model is not scalable and repeatable due to competition (contradiction)