And this is better how?
So you don't rely on a single vendor that might bite the dust, you rely on 10 of them, each one of which indivually or collectively could bite the dust. I'm assuming that even if one component of your Glueware comes unstuck that the rest is also not going to function properly. You'd need a major rewrite at that point to begin with.
Either do it ALL in-house and do it ALL yourself to make sure you are not reliant on a supplier, or go with a single supplier with contractual provisions about getting full sourcecode and knowledge in the event of a bust.
This really is one of those "solutions" that only make sense from behind a beancounters desk, after 30 conceptual meetings, 23 vendor sponsored "informational" events (preferably overseas) and a whole bucketload of useless managers have had their say.