I also agree that a lot of the benefits and crude aesthetics is lost on using extra codepages and fancy characters: I enjoy my NetHack on plain ASCII mode as it doesn't depend on any particular font for instance.
Your 'choices' are overblown. For example, in Far Cry 4 you have the option to kill a character or not (on the orders of another). In your report, you refer to them as having been "taken care of". You can't even change the dialogue. Your reward is a different epilogue movie. It's all more choose your own adventure than choose your own War and Peace. You can see the game designers dilemma - why expend massive effort on branching paths when any given player will only choose one? Far Cry 4 also had an amusing bug - you could also choose to destroy a temple or not. However the bug meant the game would show you the non-destroyed model when past a certain range, suddenly popping to the correct model as you approached. The mask literally slipping!
Even non-open world games stuggle to provide meaningful agency. The much acclaimed Walking Dead allowed many choices, from which all resulting divergences were soon neatly resolved funneling the player to exactly one ending.
Honorable mentions : Fallout New Vegas, which offered so many options it almost collapsed under its own weight, and until Until Dawn, an interactive horror movie in which every single character can live or die.
Either the author, or I, really misinterpreted this book. As I recall, it was Bembo's word processor that took a few dozen phrases—some of which were about occult sects, some of which were connectives "If this, then", and some of which were garbage "Mickey Mouse is Minnie's husband". And it didn't synthesize anything; it spat back a random selection of them, and the characters then attempted to synthesize the results. Which seems like a minor hair to split, but it really reflected one of the themes of the book, that humans keep trying to find patterns and order in the natural world, and will create it if it isn't there.
However, a similar remark is in the blurb on the back of the book, which has always annoyed me. So maybe it's me that's misremembering (or maybe the author of this article didn't really read much of Foucault's Pendulum.)
Everybody admires his extremely deep knowledge of the Italian language and its nuisances, I would even dare to say "unrivaled" in our present day.
On the other hand, certain books he wrote are considered extremely complicated to grok, or too academic.
I always found it strange that in the U.S., conversations about, say, Foucault's Pendulum, don't really touch on the fact that people are reading a translation and not the original work. The vast majority of the readers probably wouldn't even be able to tell you the name of the translator.
I always wonder how different the translation is from the original, and how different it would be from other possible translations. I remember looking at different translations of Lysistrata and feeling like they were almost completely different plays.
I do think it's funny when the translator makes a point of shitting on other translations in the foreword. I think the version of Aristotle's Politics I read in college did that.
At least for "Norwegian Wood", I found the translation to Norwegian to "feel" better than (the excerpts I've read) of the English translation. It's hard for me to say if that is because Norwegian is my first language, and I learned Japanese from living in Japan, or because there are some parallels between terse Nordic prose and poetry and Japanese. I could certainly see the Nordic sagas working "better" in Japanese than in English.
But another reason I could see to re-translate something besides previous translations being flawed is them being old. If a particular obscure work only has on English translation, and it's 100 years old, it seems pretty reasonable to want to translate it into more modern English.
Untrue, but widely circulated and accepted myth 
Some words describing how it lays, some describing how it falls, some describing what it turns into, moisture content, flake size. This is how you get to "50" (or just multiple) words for snow. English does it too. It's not just generic "snow" that gets described with 50 words... (but some are just flavor: pow, fresh whateveryougetthepoint)
tl;dr: they modify words by adding suffixes (e.g. instead of adjectives). So they have (can make) 50 words out of anything
I read Neuromancer both in English and Catalan, in there there were programs called ICE (Intrusion Countermeasures Electronics, so a firewall) and the programs to break them were called icebreakers.
The catalan translator (Joan Fontcuberta i Gel) managed to translate ICE to GEL (I don't remember what the acronym mean, but it made sense). "Gel" is "ice" in Catalan, also is one of the translator's surnames. And so we could have "trencagels" as "icebreakers" (literal translation).
So, yes, most things are not easily translatable, so when they are they stand up and show us the craftmanship of the translators.
There's a great Radiolab episode that covers exactly that: http://www.radiolab.org/story/translation/
I'm reading Foucault's Pendulum right now, and even though I'm absolutely loving it, part of me feels like I'm missing out by reading a translation. The other part of me wonders whether those who read the original work missed out by not reading this translation.
In general I often feel the same, but in this particular instance it might make you feel better to know that William Weaver's translations are widely lauded.
Eco famously said Weaver's translation of The Name of the Rose was "much better than the original". :) And I've always found it pleasing to wonder if that might not even be true. One thinks of a translation as being necessarily inferior to an original, but depending on the two writers involved, there's no ab initia reason why it might not be better (for some given definition of "better", anyway - clearer, more accessible, more profound, etc).
Particularly if you start with a book centered about plot, structure, and other "macro" properties of good literature, and apply an author that excels on the "micro" structure part.
Jorge Luis Borges, or Vladimir Nabokov, with their great command of vocabulary, come to mind. They are also great authors that dealt with several translations. Borges, particularly, stated that translations could improve the original, and modified his translations where he deemed appropriate.
A contrast is with the works of Stanislaw Lem, all the English translations were done by the same person.
Of these, Michael Kandel seems by far the best.
Check out Douglas Hofstadter's essay on just this topic.
This passage, no matter how it was translated, would inevitably be jarring in English in a way that it wouldn't in Italian or even when translated to some other language which also makes that grammatical distinction.
Not that I'm complaining: if anything, it fit perfectly in a book as meta as If on a winter's night. It certainly made my experience deeper. But it also made it different than it would have been in other languages which makes you think...
It's interesting how translators decided to handle a poem that consists of so many nonsense words that are meant to evoke specific ideas in an English-speaking reader.
Did you mean nuances? If so, this is the second humorous freudian slip I've seen on HN today....
As a book lover, I would say that any self-respecting reader would have at least heard of, if not read The Name of the Rose. Eco was been a great writer and is extremely famous. And I say this as an Indian.