In several different threads    there are various ideas for market-driven interest rates.
In several other different threads       people express a desire to have some on-chain support for peer-to-peer trading of BitAssets against BTC.
In one particularly contentious thread, I got into a heated discussion with bytemaster  about whether writing detailed specifications and analyses such as     is something that should be rewarded with I3 funds. (The outcome of that discussion was that both bytemaster and I support the idea of rewards-after-the-fact for good ideas and careful analysis. While I have received tips from several generous community members for my work -- for which I thank them -- I still have not received anything from bytemaster, which makes me question bytemaster's commitment to the rewards-after-the-fact paradigm.)
To address all three of these issues while conserving core developer time to focus on bugfixes and stability, I think I3 should post two new bounties: One for a market-driven interest rate scheme, and one for an automated peer-to-peer BTSX / BitAsset-to-BTC market (with possible BTSX chain modifications to support escrow.)
There should be a pre-specified split between the phases of specification (30%), analysis / pointing out attacks, corner cases, unintended consequences of a specification (20%), blockchain / CLI implementation (30%), and GUI / web wallet implementation (20%). In the case of the implementation phases, the bounty becomes payable when the implementation is merged into official I3 and/or dacsunlimited client. In the case of a specification, the bounty becomes payable when an implementation based on that spec is committed or merged into the official client (regardless of whether the implementation was developed by core developers or by external developers hoping for the implementation bounty). In case more than one person works on a phase and they can't decide how to split the bounty, bytemaster has the right to make final decision and the bounty payout is reduced (by 20%) to encourage the parties to make their own consensus. In the case of analysis, each individual problem should be worth ~10% of the analysis budget (so ~2% of the total bounty amount for each problem).