I do not see much of a point in seeing people's faces and with full video either resolution is low, or jittering or the badnwidth is huge. So the solution is to have a simultanous voice and shared white board, which should be controlled by the individual elctronic tablet devices (mouse is not good for drawing). The electronic pen on the tablet is giving a one-dimensional (possibly with described path width, color etc.) trace, so if one just transmits the changes with the same resolution it amounts to many orders of magnitude lower data bandwidth.

It would be nice to have (and it is easy for professionals to create such a thing) a software for recording the lectures in a format which would include just the voice and the electronic pen on tablet path with the timing information on the movement of the pen path synchronous with the voice. I do not know which voice and whiteboard formats would be based on such synchonous timing information interspersed with audio and new path information. Of course one should be able to erase, and do the recording in chunks. Application which would be a viewer for such recording would be like a video but having a far higher resolution with smaller memory used as only 1-d information is added and most of the rest of the picture is static; and the voice is not that big of a deal. This kind of setup is a must for mathematical community. Many mathematical centers do big efforts to make videos where the focusing, jittering, shade and other problems as well as the mere size of video files make it impossible to be in good resolution and not hugely expensive or of big memory size. The sound + timed chalk/pen-path solution is much cheaper and scalable to hi resolution at present level of resources and technology.