When it comes to C-band, a multifocus solution is not just an option, in most cases it's necessity. The reason is quite simple - there's no LNB that would be able to receive all C-band signals without the need to rotate the LNB or inserting/removing the dielectric into the feedhorn. For me this is the most annoying thing about C-band - of course, you'd find Ku-band examples as well but those don't come close to how much scattered the polarization is in C-band. I'm sure you've noticed that many satellites in the far Eastern and Western arc use linear polarization, while the majority of European/African ones broadcast using circular polarization. For example, to receive Intelsat 11 (43°W) Intelsat 34 (55.5°W), Intelsat 21 (58°W), Intelsat 20 (68°E) or Intelsat 17 (66°E), you'll need a linear LNB. On the other hand, SES/NSS satellites require a circular setup.
Honestly, I have no idea why such divide still exists. Both linear and circular systems are used in the same markets and the satellites provide the same footprints and signal levels - for example comparing SES-6 (40°W) with Intelsat 34 (55°W) will not reveal any advantage in using one or the other form of polarization. Circular is easier to set up in the beginning because you don't need to adjust the LNB skew in order to find a signal. But then you need to play around with the dielectric plate or a polarizer to achieve correct conversion to V/H on both polarizations, which isn't always an easy task. I've stopped using dielectric plates completely since it was impossible to find the optimal position for all transponders - I always had to make sacrifices. Instead, I'm using polarizers which seem to perform better accross the spectrum.
Unlike circular, linear signals require precise LNB skew setting but that's done just once and that's it. I haven't noticed the need to adjust the skew later in order to optimize particular frequency while sacrificing others. The signal doesn't need any polarization conversion, which just sounds better, right? Why "fix" something if it's not broken! The hard part about linear LNB setup is figuring the correct skew settings when trying to find a satellite - if the skew is not right, you won't find anything - with circular LNB, the worst that could happen is swapped polarization, although there's not really such a thing, it's just a matter of how your receiver understands R/L polarization. It's almost a standard to consider V=R and H=L.
The bad thing about all this is that in order to switch from linear to circular, either a manual intervention is needed or a robotic solution must be designed. Both of these ways of getting around the polarization mess aren't user friendly and are likely to lead to signal loss for all polarizations. That is why I have decided to use two separate LNBs with my dish - one for linear and the other for circular signals. By doing this, I achieved the reception of all visible satellites in C-band without having to adjust anything. Of course, one of the LNBs had to be pushed away from the focus. I've chosen the circular one simply because 40°W signals are stronger with more favorable FEC than linear signals at 43°W. Unfortunately this meant a loss of somewhere between 0.5-2 dB, depending on frequency, for all circular signals. But on the other hand, linear reception was optimized.
Losing 0.5-2 dB meant sacrificing some of the transponders at 40°W and everything important at 47°W (Viacom, SES UHD transponders). I'm not able to receive 3842 L (Disney), 3858 R (Sun HD) and 4065 L (Food Network HD) from 40°W and others are on the edge, such as 3980 L (ESPN Brasil) and 4131 L (City TV). Other transponders can be received well without limitations. Luckilly I have noticed no signal degradation from the linear satellites, despite shadowing part of the dish with the additional LNB's scalar ring. As you may notice from the pictures, part of the scalar ring was cut off in order to place the second LNB as close to the focus as possible. By cutting piece of the linear LNB's ring, I believe that slightly better results could be achieved from circular satellites. However, I don't want to do this as I'm afraid of losing signal on linear signals because of this and also the AvComm scalar ring is a heavy and thick beast that can't just be cut by an amateur.
In my case, the LNBs are placed approximately 7 degrees apart, which means I still need to move the dish in order to position one or the other LNB to the optimal place. But that's something I'm doing automatically at channel change and moving the dish from 40°W to 43°W, which actually means to move it by not 3 degrees but 4 degrees, takes less than 5 seconds so that's fine. And while the dish moves to the position receiving 40°W on the circular LNB, the linear one is aimed at 47°W and able to receive 4055 R (Viacom) with some nice 10 dB most of the times thanks to the fact that there's no LHCP signal on the same frequency, so the signal is degraded by 1 dB but still strong enough for a lock by a linear (Vertical in my case) LNB! That's a nice bonus. For the record, the off-focus circular LNB receives 4055 R at 47°W with only 9.2 dB, not enough to lock. So that's 9.2 dB with circular LNB off focus, 10 dB with linear LNB in focus and 11+ dB with circular LNB in focus.