I can say from experience that typical game projects start with a "blue-sky" period -- at least those not tied to a inflexible IP or publisher -- where almost any idea is entertained. However, as the project progresses through milestones, towards the inevitable final build, those ideas, which once looked so doable, transform into risky and foolhardy endeavours. So when I hear a developer has cut a feature midway through development, I'm never surprised.
Such is the case with 4A Games and Metro: Last Light, with the developer announcing that the multiplayer component of the game won't make it to release. Here's a snippet from the news post on the game's website:
Throughout the development of Metro: Last Light a small, dedicated team had been working on a number of multiplayer prototypes. After E3, we decided to fold this multiplayer team back into the main group and focus 100 per cent of the studio's resources on the single player campaign.
A small FAQ is provided after the announcement, clarifying that the multiplayer component wasn't really developed beyond those initial prototypes and that dropping it from the final game won't affect the title's 2013 release date.
The developer admits that mentioning MP early was probably a mistake, but it felt it was better to be open about it. It also didn't rule out the possibility of releasing multiplayer after the game has shipped, either in an update or standalone product.
Given the choice between half-arsed multiplayer or no multiplayer -- especially if its development compromised work on what is primarily a singleplayer title -- I know what I'd pick whether I was designer, producer, publisher or gamer.
Multiplayer Update [Official site]