> does it still earn enough through ads to be worth it
That's a great question!
So I did some googling...
And I conclude* that Xing Li can afford to buy new servers. However, his track record of investing in full-time technicians & admins to properly run the site is a whole other matter.
*Stats & math stuff comparing FFn with AO3 info on which I base my conclusion below…
"Fanfiction.net has been receiving Daily Pageviews of 1 Million and should generate an estimated Daily Advertisement Revenue of $3,081 USD and Monthly Advertisement Revenue $92,437 USD."
"Archiveofourown.org has been receiving a Daily Pageviews 609,081"http://www.freewebsitereport.org/www.archiveofourown.org
"AO3’s costs for  are projected to run up to $70,000"http://www.dailydot.com/society/ao3-fundraiser-auction-fanfic-archive/
from 2011: "our DB server was US$5,554, and the Rails twins were US$4,193 each. The storage server was US$3,014, and the switch that lets them all talk to each other was US$280. In total we spent US$17,234"https://archiveofourown.org/admin_posts/109
So if AO3's expenses ($70k) are 60% of FFn's (if we base it ONLY on pageviews, which I know isn't nuanced), then FFn's yearly expenses = $116,666.
If FFn earns $92,437/month in ad revenue (which sounds like a preposterous overestimate) then it only takes less than 2 months for FFn to pay for its expenses for a year.
4 or 5 months of FFn ad revenue sounds more plausible, and if that's the case, then yeah, Xing Li can buy new servers.
FFn's main problem appears to be the maintenance of its servers, which requires HIRING PEOPLE as regular staff. He won't even hire people to moderate site content, so I have no idea if he has any dedicated employees to care for the site, or simply calls in a contractor to fix stuff that's broken (which would, frankly, be a nightmare for the poor contractor to deal with, especially if previous contractors refuse to deal with FFn anymore and no one is around to explain a previous technical mess).