You are correct for the big boxes, but there was clustering over ethernet for the MicroVAXen.

... Nuff said.

That was voting machines with touch screens used in voting in USA.

In the Norwegian tests, the participants used private computers with a web based application.

It is not a technical issue, it is much more a political issue. Moving to an all electronic voting systyem has not even been discussed. Getting political and public acceptance for it would tale a long time.

All voters in Norway receive a voting card in the mail, and I can assure you that the norwegian mail system is very reliable as well as the cards being sent out well in advance of the election.
Regarding the learning curve... Norwegians have been able to file the tax returns electronically for a number of years, and in 2012, approx 75 % of those who filed did it electronically. Also we have one of the worlds highest use of electronic banking (I have been physically to a bank exactly once during the last ten years, and my current bank does not even have such facilities. Everything is net based).

So basically, it is reasonable to assume that the results are as presented. Turnout is not due to lack of easy access to voting facilities.

It adds complexity, as it will be in addition to the manual voting system.
There are no problems with keeping only the manual system, while it is impossible in the forseeable future to use only electronic voting.

When I vote, I pick a list for the party I vote for, and put it (unmarked) into the ballot box where it is mixed with a significant number of other similar lists. There is no way to track exactly what piece of paper I put in that box. So my vote is anonymous.

The aim of this test was to measure if there would be an increased turnout. By the design criterium, the test was no success. As I did not create the design cirteria for the test, I can hardly be blamed if the test used irrelevant criteria?

Any system relies on a certain amount of trust, but a lot more people are involved in a paper ballot. The votes are counted locally, and all the numbers are available online. The paper ballots are kept in case there is a reason for a re.count.
That means to corrupt such a system an entity would need to control a significant number of people. That is a lot more difficult than to fix a centralized electronic system.

The goal of this test was to test technology and to check if easier access to voting would increase turnout.
If you test somethig for a specific purpose, then surely accepting the outcome cannot be a problem?

As for the reason for the low turnout, that is a mixed issue. At least we can now assume that access to voting facilities is not one of the problems. As for the country in question, a few reasons may be a generally high standard of living combined with no major fundamental differences between the political blocks. (I live in that country, and my family all vote.)

Just to clear this one up: In the Norwegian tests, there were no dedicated voting machines. The voters used their own computers, voting from home. Using dedicated voting machines instead of paper was never an option.

There are no voting machines involved, as the online voting was done from the voters own PC. There is already systems in place in Norway to ensure user authentication (used for filign tax returns etc...), so any issue would be with the central systems. In its simplest form, it is a question of trust.

The problem was that the overall turnout did not increase. So 38% of those who would have voted anyway chose to do it electronically. As developing and maintaining a complex system that is used every second year would be quite expensive, along with privacy issues etc., making it a little more convenent to vote is just not a good enough reason. At least not at this time.

Norway never had general online voting. Less than 10% of the population were part of the tests.