Touch and Audition use Identification of things by parts, Vision has
both part and whole images to identify whole things.

Touch uses Active and passive Identification methods combined,  audition
and vision are both passive only.

Touch is a close proximity sensory experience, audition and vision can be
close or remote.

Touch is a one-to-one source to receiver relationship, audition and vision
are both one-to-many mappings (or broadcast).

Touch is good for examining details smaller than a person, Audition is
good for gaging distance relative to the person, and Vision is good for
gaging both distance and size.

Touch is  not as useful for recognizing affect (or emotion) of a subject,
but audition and vision are useful. (E.g. You can hear and see sadness or
happiness.)

Touch is true (or hard to fake), while audition and vision have illusory
possibilities.

Touch makes the longest lasting impression on our psychosocial development
(e.g. development of preemies), while audition and vision have lesser
effects. (From Natural History of the Senses).

Touch and audition can be used in darkness, vision needs light.
 

2.First of all, using the videolink is a must.

One way of increasing the communication could be by awareness of the
weather in each place. People often begin talking about the weather, so
perhaps a localized weather display of each other's country could help. I
like the weather display that was presented last year by Stefan Marti and
partner. It contained effects for light, wind, sound, lighting and
rain.  While this is not my design, I do believe it has the best
combination of sensory experiences for both sides. We will be able to both
hear weather sounds, smell water, reach in and feel, and perhaps even see
the time change and weather in the other place.

Another way would be to have a shared drawing surface, like Clearboard.

How about having an audio spotlight that people from each side can aim at
the people from other side to single out the person  they want to talk
to? This way the conversation can sound "intimate", and can take the place
of the normal visible and proximity clues of people passing near each
other who stopped to chat. This also avoids confusion of separating people
who want to just watch passively and people who want to talk.