Hi pepa, I set off down a similar route some time ago. I was looking to use microwave mags + neo infill for a largeish alt. I have dismantled close to 100 microwaves and recovered more bits than I'll ever use as well as getting seriously ofside with my wife due to the piles of carcases in the yard. I have since obtained a swag of nice large neos to do the job so have happily given up the microwaves. I will hopefully get around to using them sometime soon.
As to your question, instinctively I would say that using all of your magnets on one disc with laminates opposite will give you the most output, however my experience with neo infills in ceramic magnets leads me to offer caution. You have probably found that the neo magnets have no trouble remagnetizing the ceramic mags. One thing I found while testing the neo infills, was that unless they are in the appropriate magnetic circuit, the flux from the neo is large enough to re-magnetize the surrounding ceramic magnet. This is illistrated in this Vizimag simulation:
This is a cross section of a ceramic mag with a neo in the centre with no backing. You can see that the flux lines from the neo travel back through the ceramic mag in the reverse direction. I used a gauss meter (hall effect sensor) and determined that the ceramic was permanently effected as a result of this configuration (you can return the magnet to its original flux configuration by the appropriate application of a neo). In the proper magnet circuit (such as in a dual rotor setup) the relatively intense flux of the neo is given an easier return path that does not fight against the ceramic:
Here you can see that both the ceramic and neos that form one pole do infact contribute all of their flux in the right direction. I think that this (dual rotor) setup would best utilise the ceramic/neo combination, however you need to consider what damage(flux reversal) is done to the ceramic magnets as they are placed into this configuration - I think that by the time the second rotor is placed opposite the first to achieve this, the flux reversal in the ceramic has already occured... Also, if you intend to place the poles as close together as they are in the picture, I think that the contribution of the ceramics will suffer even further due to leakage.
With so many factors unknown, I feel that the best way forward would be to wind a coil and test the options. You might consider using the tops (the 'I') of the microwave transformers for the laminated backing for such a test. I find the easiest way to separate them is to make a single cut along the weld approx 2-3mm deep using a 1mm cutting disc on the grinder, then give them a tap on the side with a hammer - should come right off. A couple of these should be fairly effective for testing a single coil setup.