syn_weight semantics

[arsenius7]

According to the current spec,

syn_weight - Strength of the connection between the source and target nodes. The units depend on the requirements of the target mechanism. For example, if the target mechanism is NEURON's Exp2Syn, then the syn_weight is interpreted as a peak conductance measured in microSiemens (uS)

Imho, reusing same property name with different semantics can easily lead to misintepretations (at least, from less simulation-savvy people, but the format should be defensive against this type of users).

It should be either given different names depending on the model used (say, conductance for Exp2Syn vs syn_weight for instanteneous), or at least have an attribute with corresponding units (uS vs ?); ideally both.

[arsenius7]

Having units would have greatly reduced the chance for misinterpretation.
Closing the issue in favour of #66.