Multiplayer won't necessarily take resources away from other features. There's no confirmation here that multiplayer will sacrifice anything from singleplayer, people are just reading that into it. Imagine simply adding the ability to co-op or join privately hosted servers, without doing anything else to the game as far as balance or content.
Implamenting multiplayer would take developer resources, a
lot of developer resources, and these then become developer resouces that can't be spent on other features.
First, you need the basic multiplayer code. I don't know how difficult this is, but it's still work that needs to be done.
Second, you need the rest of the engine to support multiplayer. This is difficult, and in the case of the Elder Scrolls it prevents Bethesda from streamlineing the game engine for single player. Currently the game only ever needs to worry about one player, the scripting system only ever has to worry about one player, and objects don't need to be tracked across multiple computers. Terrain, NPCs, and objects also only need to be loaded around one player, and if the system needs to juggle more balls it would really ramp up the hardware requirements.
Third, you need this to work smoothly. Multiplayer isn't something you slap on and it works great, and I suspect in a game like Skyrim it's particularly difficult to get everything running smoothly.
And finally you have the potential for design and balancing issues. Are locations large enough for four people to run around them? How should difficulty be scaled? How does the AI deal with the increased difficulty of two or more fully developed characters? Some of this needs to be considered anyway for companions, but they aren't nearly as formidable as a player both in terms of stats and intelligence.
As I mention in the opening post it's a real shame I can no longer find the Soldier of Fortune 2 portmortem online because it does an excellent job of discussing just how time-consuming it was for them to implament multiplayer. And this was with a game engine specifically built for multiplayer, so they didn't even have most of the hurdles Bethesda would. If Bethesda wanted their next engine to include multiplayer they'd have to put a fair amount of resources towards this, which would mean those resouces aren't being spent on other features. Perhaps if Skyrim had multiplayer we wouldn't have radiant quests. Or maybe they would have had to hire more programmers at the expense of designers or artists.
Now Bethesda may decide one day that this is worth pursuing. However I've been following their comments on this matter for a while and they've made it very clear, in my opinion, that they aren't interested in dividing their focus. If we do see another multiplayer Elder Scrolls game I suspect it will be a spin-off developed by one of their sister companies, and not part of the core series. (Note, there's precedence for this. The last multiplayer Elder Scrolls title was Shadowkey, developed by Vir2L.)