Two key differences between these two forms of communication were publicity and social cue transmission.
Communicating over ICOM was awkward because it was not apparent who else was in the room that was not in the field of view of the camera. Also, even with the use of headphones, the conversation becomes public, for the voice is heard (and sometimes amplified to make sure it gets across) in the space. The location of the ICOM system, makes for a rather public conversation.
In contrast, the face-to-face meeting was more private even though it was at an Au Bon Pain. It was easier to control the volume of the voice and there is this expectation or assumption that each table is a private sphere in this public space.
Cues were a very obvious difference. Using ICOM gestures were exaggerated, there were delays due to ambiguity of social cues and they oftentimes had to be reenforced by speech. This was most evident in the conclusion on the meeting where we explicitly said it was the end. This is more obvious in physical spaces with gaze, looking at a watch, or simply closing a notebook.