I don't see how dictitat can be a(n) historical present when viewed in context. In Verrem 1.1.4. It seems to me that Cicero is talking about what Verres is going around saying right now--and Cicero is appalled by it. He's emphasizing the resources that Verres can bring to bear against prosecution, and Verres' attitude that he's above the law. An historical present would be appropriate in a narrative of a series of past events, which this isn't. You can't just drop a single historical present in a passage about things happening in the present and expect the reader to understand that you're suddenly talking about the past. And I come back to "why sit?"
I also can't see this as what Woodcock calls a "generalizing present" with such words as aiunt, dicunt, ferunt, traditur (see Woodcock, sec. 279(d), p. 233). These are words such as "they say", "it is said" "it is reported", with an impersonal subject. Here Cicero has Verres talking.
And I think that if dictitat were intended to convey a suggestion of repeated utterances in the past as well as the present, that would have been made clearer. Again, why sit?
But Willcock (p. 234) has something interesting to say after trying to explain away other apparent exceptions to the rules of sequence of tenses (none of which seems to apply here):
"There remain isolated examples which it is more difficult to explain on the above lines, e.g., Cic. de Rep. 2, 30 multa intelleges etiam aliunde sumpta meliora apud nos multo esse facta, quam ibi fuissent, unde huc tralata essent. One would naturally suppose that fuisssent and tralata essent stand for O.R. fuerant and tralata sunt. If so, then it must be confessed that even Cicero could be inconsistent in his syntax . . . ."
He goes on to offer an alternative explanation (a kind of conditional or "past potential"), but he doesn't seem entirely comfortable with it.
A. Ernout & F. Thomas, Syntaxe latine, sec. 399, p. 411, in a discussion of exceptions from the sequence of tenses rules, give a number of examples where subordinate clauses in indirect discourse have "librement (freely) the tense voulu par le sens (intended by the meaning) whether or not it is in concord with the principal clause." There are four examples from Cicero of instances where the imperfect or pluperfect subjunctive occurs in subordinate clauses in indirect discourse after a main verb in the present tense.
It's worth bearing in mind that the rules about sequence of tenses were constructed by grammarians on the basis of a large corpus of Latin texts (especially Cicero). Isn't it to be expected that once in a while, a good writer of Latin -- in fact, the supreme master of Latin prose -- occasionally slips in a usage that doesn't conform exactly to the patterns that emerge from the corpus?
It's still puzzling that (1) as you note, the commentators haven't seen fit to explain the apparent anomaly, and (2) there doesn't seem to be any special nuance associated with what is apparently a departure from the usual rules of sequence of tenses (thanks for setting me straight on the rules).
However, maybe by dismissing other officials' transgressions into the past sequence of tenses and putting Verres' depredations into the present sequence--in a slight bending of the grammatical rules--Cicero's Verres throws into dramatic contrast their small-scale pilfering and his own transcendent rapacity.
You were very astute to notice the anomaly, just like the one in the passage from the Phaedo.