Your basic design options on the dc side of the system are described in Diagram 2a & 2b (which uses a Fronius product in the example):
http://solarprofessional.com/article/?file=SP2_3_pg28_Jacoby&search=
as are some basic design considerations:
* What is the cost difference between a transition box?plus the necessary internal components?and a combiner box?
* What is the cost difference between installing multiple PV source circuits or one PV output circuit?
* Will using a separate combiner box require a separate dc disconnect switch? If yes, what is the additional cost to install?
* Will using a separate combiner box mean that the PV output circuit current rating exceeds the input rating of the inverter input combiner terminals or fuses?
* Does bypassing the integrated fusible combiner violate the product listing?
* Can the combined output circuit be installed per the manufacturer?s specifications, as required in NEC 110.3(B)?
* Will combining the conductors change the required conduit size? If yes, how will this affect material and labor costs?
* Are there any other benefits to combining? Will the wire pull be easier or harder? Will fewer or more dc disconnects be required by the AHJ or the Code?
Your situation is complicated by the fact that you are transitioning conductors off of wo separate roof surfaces, but the basic thought process is similar.
The logic behind Code-compliant conductor sizing is discussed in this article, which also includes examples:
http://solarprofessional.com/article/?file=SP4_3_pg14_QA_1&search=
RE: the AC OCPD. Yes, the max continuous output current from the inverter (47.5 A @ 240 V, 1-P) is intentionally set to let you get the most PV capacity possible on a 60 A OCPD (47.5 A x 1.25 = 59.4 A, which is less than or equal to 60 A).