...

Our current igniters work just fine, but they could probably be better. We could look into using nichrome bridgewires, a different combination of our current ingredients (quickburst, chunks of propellant, propellant dust), or a new pyrogen mixture like MTV (magnesium, teflon, viton), or BKNO3 pellets. Lots of quickly testable stuff here, but it all HAS to be done in the blast chamber for numerous reasons.

...

Attach the propellant directly to the inside of the case rather than using casting tubes and a liner. Provides better performance by increasing case volume loading. Requires new adhesive strategies and necessitates a monolithic design, which makes casting a challenge. 

Better binder understanding

Develop a better understanding of what our binder is like. Do a mix with only liquids and use this result to calculate theoretical density more accurately. Monitor the batch frequently during cure to figure out how long we actually need to wait to demandrel. Experiment with removing/adding some of our processing agents, because we probably aren't doing it right. Figure out if expensive HTPB with HX-752 is essential or important at all. Maybe cheap stuff with tepanol or no extra additives would work just fine.