I want to make a comment about Dash needing to act like a "real" company.  This isn't the goal that we want.

A "real" company has a structure that is typically centralized and managed from the top down.  This requires trust from all parties.  You trust the CEO won't run away with the company earnings, the middle manager won't spend expenses on a new sports car, and an employee won't send that secret design to another company.  This is controlled somewhat by laws(stealing funds, company secrets) etc. The threat of losing a paycheck is also very real for the employees or middle managers.  The top managers usually own part of the company and don't want it to fail or they loose their investment.

So now we try to fit this into a dash budget system.  It doesn't work.  Dash is decentralized - not centralized and top dog owned.  So if we assign project managers with decision making ability onto a class of people that don't have an investment nor have specific knowledge for doing that project - we lose the incentive and skills to be efficient.  The only risk they have is losing their position and not getting paid next payroll.  This is a bad interpretation of fitting a decentralized organization into a centralized model.  The 'control' that the core has now, will probably not last in this model either.  If a bad decision gets made, it is very likely masternodes will vote out the entire core team.  The new team that replaces them could be much worse.

The interesting thing about decentralization is that it isn't controlled.  There are a lot of what seem to be random events that happen.  Sometimes there is waste from duplicate effort.  Sometimes there is quick idea generation and execution.  That doesn't mean that it is bad, it is actually much better than centralized for new ideas.  Evan suggested that developing 2 Android wallets was a bad thing - Bitcoin has many mobile wallets and it is great to have options.  At some point, when we have 2 or 3 wallets(or even one wallet with all the feature people want) then maybe masternodes will consider not voting for new ones.  But just because 2 similar things are developed it is not necessarily wasteful.  Many companies develop 2 versions of a product to see which one does better.  Then they only release the better one.  Right now we have some big core and awareness proposals that don't really have any specific goals.  If we can tie all proposals to actual activities, then voting each activity won't be a guessing game in assumptions for what the 'big bucket' budgets will be covering.  And duplication and priority of activities will be easily decided by the masternodes voting.

Let's keep this decentralized budget model going with maybe a few tweaks to make voting out a contract a little harder and I think we have what we need.