Through the discussion, he gives special place to the dative agents used with perfect passives and their decline. George suggests that the agent dative with perfect passives was a result of the semantic nature of the perfect as stative. This is from his concluding summary:
George also notes that even in the Koine/Hellenistic period, the dative agent with the perfect passive does not actually change its distribution. The same two environments still hold, but the number of perfect passives with overt agents decreases dramatically. With regard to Plutarch, George makes the following statement:The perfect passive ... did not like the present and aorist passive, describe a dynamic action ("The door was opened"), but rather an unchanging state ("The door was open"). It was thus incompatible with an agent marking like ὐπό. Later, by the fifth century BC, as the transitive perfect active became more widespread, the perfect passive, increasingly viewed as its intransitive counterpart, came into closer alignment with the present and aorist passive. ... At this point it was longer purely stative. ... But ὐπό at first only occurred with the perfect passive in two environments. First it was used when a noun in the dative might have been interpreted as something other than an agent [MGA: e.g. an indirect object]. Second, it was used when then [subject] patient of the verb was animate, e.g. "The man has been sent by the king." (266-7)
If this is an accurate description of the development of this Greek verbal form from denoting stativity early on and then developing into a "standard" perfect, then I would wonder if we could extend George's idea for the dative of agent of the perfect passive in the Koine period as a higher register to the use of the perfect passive itself with certain verbs. Specifically, I'm wondering whether the extremely common use of γράφω in the perfect passive in the NT. If this is more of a high register archaism used centrally in quite formal contexts, perhaps the perfect passive γέγραπται (and other similarly formal perfect passives) should not be taken as any kind of representative use of the Koine perfect for normal speech. It seems quite clear to me that γέγραπται is one of a number of verbs that continues to quite clearly be stative in its meaning, but perhaps that has little to do with the function of the perfect, as a whole, during the period and merely a formal language relic of the past similar to the use of thee and thou in Modern English*.As for the two examples from Plutarch, one occurs close to the beginning of Phocion, in a context that suggests that the dative of agent, by now rare, could be used as a higher register rhetorical device:
(14) Plut. Phoc. 1.2 τοῦτο δ' εἰ καὶ τῷ ῥήτορι θρασύτερον εἴρηται
"and if this has been said by the speaker rather boldy"
*It should be emphasized that this conclusion in no way suggests that the perfect passive was always formal. To extend the analogy of English pronouns, thee and thou, historically, were previously the common pronouns that took on a formal register after they fell out of use in normal language.