An ideal parabolic reflector can only bring all of an incoming signal to a single focus point if the source is on axis. In the case of multi-satellite reception, any off-axis source will suffer from coma distortion, which will spread the 'focus' away from the apparent, off-axis reflection. One can model this numerically to find the best locations to place feeds, but there are a lot of factors (signal frequency, dish and feed dimensions, location on the earth, and the specific orbital locations) that play a role. I wrote some software to do this, and it accurately predicts where to place feeds on my dishes, most of which have multiple feeds. This is heavy lifting territory.
Of course one can simply center a dish with a single feed, and start hunting around for the off-axis locations. This is really hit-or-miss, because you can receive an off-axis source in substantial range of locations, but you want to find the one with the smallest 'circle of confusion'. That can be rather counter-intuitive, particularly for sources far off-axis. You can probably get something to work, but in many cases it will not be optimal.
After modeling all sorts of dishes (I have a lot), I can offer several generalizations. Choose a big dish with a high f/D, preferably around 0.4 or higher. This will spread the signals out at the feeds, and you will experience smaller losses off-axis. Something like a Winegard Pinnacle is the polar opposite of what you want, because it has a low f/D (0.275).
As a contrived example, let's say you want to use a quad feed on a 10' Pinnacle to get 4 orbital locations, each separated by 2 degrees. It turns out you can't get C-band feeds close enough to do this, but even if you could, the two inner feeds would have nearly 3.5 dB of CNR loss compared to a single, on-axis feed. This is about the same gain as a 6' dish. The outer feeds would lose nearly 5 dB. You're not going to get anywhere close to 20 degrees at that rate.
Now take a f/D = 0.4, 10' dish. The inner two feeds can be separated sufficiently, and will have a negligible loss compared to a single on-axis feed. The outer feeds will have to be squeezed as they are just slightly too large in diameter, but you'll lose less than 1 dB, roughly equivalent to a 9' dish. As a practical example, I have a feed offset by 8 degrees from center on a f/D = 0.38, 12' dish, and it performs like it would on a 8.5' dish. Not incredible, but usable.
The problem I have with dual and quad plates is they are unlikely to hold feeds where you need them. In most cases the problem is the feed diameters constrain how close one needs to get them spaced, and creative means are needed that cannot work with these plates. But if you end up with a dish 14' or greater, they might actually work.