Surely, the claim that a method/theory can produce results possessing properties that the method/theory does not postulate/support would be dismissed as a proposition to dabble in ALCHEMY.

This is why it is so surprising that propositions of this nature are made by senior scientists in the peer-reviewed literature!

One example should suffice.

A number of senior risk analysts/scholars propose in their writtings that decisions maximizing the likelihood of an acceptable outcome can be yielded by a decision model that is based on a likelihood-free model of uncertainty!

When challenged to justify such claims, namely when asked:

How can a likelihood-free decision theory possibly seek, let alone yield, decisions that maximize the likelihood of an acceptable outcome?

they argue:

Well … perhaps our phrasing was not right, but you know what we mean!

To which all one can say is the following:

Indeed we do, but … in this case your claim can only be described as a proposition to practice Alchemy!