Before really getting to this we had a good discussion on the usefulness of the concept of tangible interaction - obviously we see the advantages clearly - but nevertheless it seem in many ways hard to proof. The argument for tangible UIs as manipulators and controls is very clear and can be shown but looking at tangible objects as carriers for data it becomes more difficult. Looking a physical money the tangible features are clear and one can argue for the benefit of tangible qualities (e.g. I like Reto's statement "the current crisis would not have happened if people would have had to move money physically") - but also the limitations are there and modern world with only tangible money would be unimaginable.
Taking the example of money (coins and bills) two requirements for tangible objects that embody information become clear:
- The semantic of the information carried by the object has to be universally accepted
- Means for processing (e.g. reading) the tangible objects have to be ubiquitously available
 Dermot McGrath. No Pain, No Game. Wired Magazin 07/2002.
 Shin'ichi Konomi, Christian Müller-Tomfelde, Norbert A. Streitz: Passage: Physical Transportation of Digital Information in Cooperative Buildings. Cooperative Buildings. Integrating Information, Organizations and Architecture. CoBuild 1999. Springer LNCS 1670. pp. 45-54.