The answer is 'no', essentially always. Spam is one of the worst abuses of the system and we want to prune it out on the spot. The usual recourse for a site moderator is to destroy the account immediately with a few quick clicks. Applying this option, the software in place has various mechanisms which over time make it harder and harder for a spammer to return to ply his trade.
Removing a spam link only hides the abuse and complicates site moderation.
It must be exceedingly rare that a spam post actually does have content that meaningfully answers the question. Many spammers try to conceal a spam link within superficially relevant content, typically composed by a quick copy-and-paste from a Wikipedia page, and usually it will be obvious to all what's going on there. In any case, wherever you see spam in the sense explained above, the best option to raise a spam flag right away and let a site moderator destroy the offending account. (I don't think I've ever seen a case where there was content really worth preserving, but what I would likely do in that case is post that content separately [as a CW post]. Feel free to tell us in the flag if you think there is content worth preserving under a new post.)
I'll take this opportunity to return to the issue of defining 'spam'. Some users interpret the word more broadly than the original sense of unwanted vendors, and use it for flagging posts which are unwanted or irrelevant for some other reason they have in mind. This would not be a problem, except that when you choose the spam option for flagging, you are partly training the software to recognize what 'spam' means, and using the word in the broader sense is then counterproductive. (I learned this from the MO user formerly known as 'quid'.)
Regarding Stefan's question in a comment about "who spams" (are they organized groups or individuals?), one of the SE Community Managers wrote to me as follows:
On the question of who spams, it seems to be a little of each. We certainly have seen waves of similar spam randomly posted around the network. A few years ago we had a bunch of posts advertising a site to stream NFL games. It seems as if someone was hiring workers from countries with low labor costs to post URLs. These were removed pretty quickly, but we still have a bunch of football team names in our list of spammy words. ;-)
We sometimes see spam from individuals: consultants advertising their services, overly enthusiastic users of a product, salespeople, etc. These aren't very organized, as a rule. Often the best thing to do is delete the content and leave a mod message. As you said on meta, any content is rarely worth keeping.
There's not a lot of bot spam that gets through. Presumably, nobody has invested much time in creating bots for our Q&A format because spam doesn't stick around very long, so it's not profitable. I imagine CAPTCHAs take care of most the rest. Our SREs are really great, so there might be some other magic on the network level I don't know about.
More information is here: What are the “spam” and “rude or abusive” (offensive) flags, and how do they work?
