In theory this is what governance tracks are supposed to solve, with conviction voting you can delegate votes to a particular individual for a particular track. The issue is how we organize the tracks, i.e. by amounts instead of topics. If we have an “events spending track” for example, you can delegate to the expert you know in the matter to vote only for events related proposals, to keep the ability to configure tracks based on requested amount then you need to combine and have different spending tracks per topic, it feels a bit chaotic so UIs should expose it nicely.
What remains unsolved is budgeting, I like the recent topic by Shawn about using the wisdom of the crowd to price proposals(we’ll implement it with our payments pallet), I would use it more for the budgeting of the main treasury and have the community decide every X amount of time how income should be distributed to different sub-treasuries(sub-daos) governed with their own set of tracks, within the sub-dao you would use the conviction and delegation system as usual to delegate to your favorite expert.
Most of this stuff is already possible in the tokenless parachain Kreivo(or will be soon). Going more to the implementation detail, we are going to add extra unconventional ways to vote on proposals beyond the binary aye/nay system and integrate that with pallet-referenda(what drives OpenGov that we extended to allow for dynamically created DAOs), for example we already have 1 member=1 vote system, vote by ranks, by other assets, custom vote multipliers that can be defined by the DAO with a smart contract or relevant to this topic one system that will give voters the ability to vote on a set of options splitting their vote with percentages or finding a median.
If people want to experiment with new decision mechanisms or have ideas for different ways to improve OpenGov ping us on Matrix 
, flexible governance for DAOs is a big part of the product we are trying to build and some of these experiments could make it into the PolkadotSDK 