This is not an accurate estimation because not all of the old customers were on the same $10/month plan, and not all of the 24 million (forecasted) customers are on the $16/month plan.
It's not accurate, but think about it: 25M to 24M is only a 4% decrease; $10 to $16 for the dvd+streaming plan is a 60% increase, and even if that translates to only (say) a 6% increase in average revenue per customer, they're safe.
Netflix's own document lays out the forecast pretty well: http://files.shareholder.com/downloads/NFLX/1403529688x0x500... (see Venn diagram). Too bad the CNN article doesn't delve into these details.