That depends on what you mean by "tedious frequency response analysis". If you mean without having to derive state space averaged models, you have a couple of choices. You could try a comercial program specializing in SMPs, like Simplis for example (
http://www.catena.uk.com/site/products/simplis.htm) but your schematic capture tool and libraries may or may not interface with it. You could try using SpectreRF's PAC analysis or perhaps a similar analysis of a harmonic balance simulator. Or you could try deducing the AC response from a brute force simulation of an impulse or step response.
However, I suspect the alternatives to state space averaging will require every bit as much work, if not more...at least the first time through anyway. You may still have to resort to state space averaging to check the new approach.
Any way you look at it, one way to simplify the analysis is to use a single converter model, scaled to represent the paralleled converters. As I recall doing this some years ago, as long as the stabilizing ramp is adequate there is no danger of unobservable RHP poles associated with oscillations between paralleled stages.