Great question, I'd love to hear input on this as well.
We split patterns into design and application patterns.
For the patterns we covered in the guidance, many were motivated by maximizing unit testing, although they patterns are valuable as well for layering/separating concerns. The two primary patterns that fall into this category are service locator and
Repository is a pattern I'd encourage many to consider. It centralizes list access logic - and most performance bugs relate to list access, so reducing the amount of code you have related to list access is a good thing. Also helps with centralizing
policies, like caching. We show both list repositories and respositories for LOB data.
These are higher level than design, and show how to solve a business problem in a reusable way using SharePoint. We show three different app patterns for event driven site creation, site instance resolution, and event coordination (this is only a miminimal
THe scenario for these patterns was creating a site for collaboration on a business event (for example, a tier 3 incident, or an order exception).
I'm looking for input on the common, non-trivial customization scenarios that are often repeated. Please feel free to email me any input for consideration in our next release! (email@example.com).