Question Max pitch&roll angles reached (without using hamster ball...)

For a classical pivot based system the theoretical max is close to 90 degrees either way, depending on the possible mechanical binding of the coupling, so getting near 180 degrees, but other factors come into play such as the mechanical efficiency of levers, which drops off rapidly over about 38 degrees up an down, it is possible to drive a rig at higher angles it just isn't very efficient and puts a lot of strain on motors, which in turn should be bigger and likely more costly.

Some of the 6DOF members may like to comment on the extreme envelope of their various designs.

In my mind this is an interesting theoretical question, but in practice extreme movement of a few more degrees is far less important that sound cost efficient design and other possibilities, such as incorporating driven or parasitic G-Systems to better provide cues for sustained Gs, which will always be beyond the performance envelope of a traditional 6DOF.

A G-System is driven via hardware and game telemetry, like a traditional DOF axis, while parasitic G-System rely on the existing movement of the rig to apply additional forces in different ways as a means to provide additional cues the rig is otherwise unable to give.

The simplest and most common parasitic G-System is members adding a race harness to say a 2-3DOF, where the shoulder straps are connected to the bottom frame and the pitch forward of the rig means the user experiences a tightening of the race harness under pitch forward, with pitch commonly being used to simulate surge on those 2-3DOF rigs, though there are many other creative and practical possibilities for parasitic G-Systems.

I would add that @SeatTime's current large heave axis rig is one example, particularly for flight simming, where a particular large axis is being preferred to already tested alternatives such as active and parasitic G-Systems, though as I understand it @SeatTime still plans to re-introduce those based on his extensive experience and experimentation with such G-Systems into his current 3DOF design.