This is pretty much it, well expressed. If we want to get numerical about it, and I might do the math but the value my time for zero reward says probably not........ starter FIP increases from inning to inning. Strictly on that basis to win more games (minimize team FIP) there is an average optimum where the hook should occur. Then, furthering the math, since a starter's early innings FIP are lower then the later innings, the more games they start, the more often you gain the value of these low FIP early innings. The numbers can be crunched to see which is expected to give a lower team FIP, 4 man or 5 man rotation.
And it wouldn't be a strict 4 man rotation. For the Jays for example, the regular rotation would be Stro, Sanchez, Happ, Cobb (?) then you have Estrada in the pen who can spot start and Biagini in Buffalo available. So its a more fluid approach, and gaining the advantage of more starts from your high value pitchers.
Injury risk a separate topic entirely. I can see a argument that pitching with shorter rest and shorter pitch count (4 man) has less injury risk then longer rest and longer pitch count (5 man).