The problem with a naive interpretation of Chesterton's fence is that it encourages stasis in cases where we don't have the knowledge or resources necessary to "fully" understand the reasons (or whether any even exist) for the current state of affairs. And in fact, there's no way to guarantee that we "fully" understand all reasons for the current state of affairs. Thus Chesterton's fence essentially deteriorates into the precautionary principle, which is bad epistemology. This quote from Wikipedia's section on criticism of the precautionary principle applies just as well to Chesterton's fence:

"of the two available interpretations of the principle, neither are plausible: weak formulations (which hold that precaution in the face of uncertain harms is permissible) are trivial, while strong formulations (which hold that precaution in the face of uncertain harms is required) are incoherent." [0]

[0] https://en.wikipedia.org/wiki/Precautionary_principle#The_pr...