hi bert, i looked this up in Devine and Stephens on greek prosody.
they say that there's evidence that greek questions did differ in pitch from statements and commands, but they say the difference is in the pitch of the question marker-word, if there is one (so e.g. [face=SPIonic]ti/j[/face] never takes the grave), but they don't say that the "semi-colon" question mark by itself raises the pitch at the end of a clause.
