This particularly makes sense for SaaS offerings that are being discontinued. If your product is only available as a service then it is particularly painful when it disappears for those who depend on it. From a business standpoint, open-sourcing is a good way to maintain goodwill and shift the onus of administering the service to those who consume it without leaving them out to dry. All your consumers will have to make a decision on whether the cost of running their own service is worth it, which is a much healthier situation than a bunch of irate customers left with no recourse.
This is not only good for your customers, but also puts some positive pressure on your competitors. Now they have to compete with an open source tool, so whatever they do, they'll have to provide a service which is at least as good as that. In other words, they'll have to compete more on quality and less on price, which is usually a good thing.
If it's truly open source then there's nothing stopping another business from coming in and offering a hosted version much in the same way that the original did.
I suspect that the reason this doesn't have so much could be licensing of other libraries etc used by the code.
If it were that easy to run a business in that market, the original probably wouldn't have went under. Companies go out of business for a number of reasons, but I'd bet on the initial founders who had the vision to create the product and have experience in that market over someone who saw free code online and decided to have a go of it. The reason we almost never see it happen (at least successfully) is that there's almost no barrier to entry (anyone else could pick up the same code) and that it's not an easy market (the original company went out of business).
