Semantically, there's a marked difference between, say, A. and B.
A. To be perfectly clear (linking verb + adjective)
B. To be perfectly understood (verb + participle)
In A. the adjective
is not part of the verb proper ([ . . . ]), whereas in B. the participle
is part of the verb proper:
A. [to be] perfectly clear
B. [to be perfectly understood]
Now the verb proper happens to be an infinitive, and its structural integrity
hasn't been violated in A., and it hasn't been violated in B. The adverb doesn't come between the infinitive marker (to) and its base verb (be). If it did, we'd be looking at a split infinitive.
With our examples, the adverb is outside the verb proper (A.) and inside the verb proper (B.), but not inside the infinitive itself:
A.
[ to be
] perfectly clear
B.
[ to be [perfectly understood]
]
If you want to agrue that B. is an example of a split, I'd be the first to agree with you, but you'll have to use a different term because "split
infinitive" B. is not.
Now if we moved the adverb out of the verb proper (
[ . . .
]), then it would serve to modify the verb proper:
C.
[ to be understood
] perfectly
With structural change comes a change in meaning, so C. and B. don't express exactly the same thing:
B.
[ to be [perfectly understood]
]
C.
[ to be understood
] perfectly
In short, B. is not a split infinitive; no "infinitive" has been split.