This site seems to share the common theme - a Braille tablet:
To some extent, the current technology, which is called a piezoelectric display is like a lithium ion battery. Its not like we don't have other battery technologies, it is just that lithium is enough better than the others, that no other can take hold. The piezo displays are almost silent, last a LONG time (20 years), and take very little power. They are, however, slow and bulky and not practical for mobile applications. The mechanism is so large, you cannot create a multi-row display. (edit: they are also rediculously expensive!)
There have been multirow displays using solenoid actuators. These were hand made by grad students and weighed a ton. They used a huge amount of power and got jammed easilly. Other than that, there are pneumatic displays with huge potential, but are also bulky. There are static electric displays that have life expectancy issues, and a few other technologies, like using memory springs, that tend to have speed issues. Then there are a huge number of mechanical "conveyor belt displays" which are too noisy to use in a classroom or office.
But all of the displays you see being publisized are not the ones that I am talking about. Displays like the one you linked to are totally 100% FAKE!!!
https://nfb.org/images/nfb/publications/bm/bm00/bm0001/bm000...
Impractical: you can stack piezo-electric Braille cells, but your device will be about half an inch high per line of text. Weight also goes up, of course.
Not affordable: ballpark, a _cheap_ piezoelectric Braille cell costs at least $25 per character (http://www.aph.org/orbit-reader-20/ uses a different technology, but is considered cheap at $500 for 20 characters). That's $3.50 per 'bit'. At that price, two 80-character lines would cost you $4000 (In reality, that probably buys you _one_ 80-character line)