Methodology and the Broken Window Fallacy

The mainstream economist accepts empirical analysis as the appropriate method to understand economics. The Austrian economist accepts a priori analysis.

Take the broken window fallacy: Does the mainstream economist accept this as a fallacy, or does his empirical analysis lead him to believe that jobs are actually created by broken windows?

If he accepts the fallacy, then how does his empirical analysis help him prove the “unseen”? He can’t prepare a statistical survey, or run simultaneous equations to measure the “unseen”, so how can he accept the fallacy? It betrays his methodology.

If he rejects this as a fallacy, and concludes empirically that broken windows create jobs (wealth), then how does he dispute the logic presented by the “unseen”?

The simple broken window fallacy is a dilemma for the mainstream economist, is it not?

Although I don’t think you can prove a theory empirically, you can at least test the broken window fallacy empirically and know that it is not false. You can empirically see that the shop owner has one window for the price of two, instead of two for the price of two.