If you want large companies to innovate (and I do) you can't complain bitterly if they fail and retire a product or service in this way (which as others have noted included a lot of notice) The flip side is big companies won't innovate or release anything new. "No way! Have you seen how people complain if we have to close it down? Not worth taking the chance of all that bitching"
P.S It's kind of funny that one of the alternatives they pushed at the time of the initial announcement was Braintree who then went on to be acquired by PayPal. Shows you how fluid/active this space is.
So why not do it faster to avoid people relying on the service for many years? Set a private internal deadline of 6 months or a year, then if it's not as successful as your initial goal, kill it then, not seven years later, after tons of people have come to rely on it.
Partly because it is hard to operate in an economy that is far too nimble. One of the reasons that it is hard to sell to a large businesses as a small business is that it is quite likely that you will go out of business - and that will cost them a lot of time and money to reconfigure if that happens.
Most components designed to be included in a larger product for sale will be available from another manufacturer, aka "second source". This is in case the first company goes out of business. Unfortunately, this practice does not seem to have made it to the software world, and so when a company goes out of business, your existing working solution no longer exists, and you will have to scramble to find another one (or go out of business yourself, if you depended on that functionality).
P.S It's kind of funny that one of the alternatives they pushed at the time of the initial announcement was Braintree who then went on to be acquired by PayPal. Shows you how fluid/active this space is.