I think it is far more important for Lisk to spend some serious time examining the entire design of their protocol, make some necessary changes, fully document everything and then have a long conversation with the DApp community about the types of applications they want to write and where the experience is lacking.
I think this process can be started and finished by the end of 2016. With respect to the next step of encouraging application development, I'd like to see Lisk become a marketplace for DApp services with delegates as the service providers. The idea would be to augment existing infrastructure and software to enjoy a newly federated or decentralized component. For example, user accounts could be stored encrypted with hashed lookups on a sidechain secured by delegates. This would give the application a tamper resistant, but secure user DB.
These kind of hybrid experiences seem to be natural fit for where Lisk could position itself. Basically it would be saying let's make all this really cool tech in the blockchain space available in a federated flavor to application developers to be consumed in modular bites with marketplace mechanics backing it.