>fabrications are usually discovered.

I think you are being far too optimistic, here. It depends on your admittance threshold for "fabrication" but intentionally skewed interpretation of data seems to be the bread and butter of academic science in many fields.