The conclusion mentioned some interesting truisms about prototyping, team-size-induced-complexity and "long bugs" vs "short bugs". Good stuff. If the PDF was not a collection of images of the text, i'd have copy-pasted that bit here. :)
The paper is dated 94 which is the same year as the original implementation as the C++ STL (https://en.wikipedia.org/wiki/Standard_Template_Library#Impl...). So it's a fair bet they didn't have access to that. I would think the entire difficulty of the exercise would change with that.
I don't think you can compare the changes between Ada 83/95/2005/2012 to what has happened to C++. The Ada language seems much more stable and backwards compatible to me, and given it's nature, it makes sense to avoid introducing major changes. However, I'm not an Ada expert so correct me if I'm wrong.
If it was written before 1998 I expect you'd find a version of C++ that existed before 1998 :)
If it was written today, I don't know that's a really good question. I think its a mistake to ignore the advances that 2011 and 2014 bring to the table though.
I do not think that someone wants to use C++11/14 to make app prototype. There are a lot more languages now which in my opinion are better suited to prototyping than languages mentioned in article. Most of people will use Python, JS or some other scripting language.
Likewise C++ isn't going to be as concise as Haskell ever.
So the paper still hold some truth if not mostly valid.