Honestly? All I see are inelegant hacks tweaking a flawed foundation.
First, I think it is a mistake to "burn" the subsidy. The way I look at it is that the DNS DAC currently has defined the total supply to be 10 billion with about 5 billion currently in a reserve fund held by the DAC. I suppose you can move (or "burn") the money by moving some of it from the reserve fund to the untouchable fund, thereby providing investors with a monotonically decreasing total maximum supply (assuming no future hard forks changing the rules), but I think that entire concept is stupid and misguided especially in this early stage when we would like to have as much of the hard-limit supply available to grow the DAC and reach its potential. In other words I think the subsidy_burn_rate variable should be zero and any parts of the transaction fees or subsidy that aren't used to pay the delegates (determined from pay_rate and subsidy_rate) should be recycled back into the reserve fund. Then, much later in the future, we can hard-fork if desired to reduce the maximum supply limit to less than 10 billion if there really isn't anything valuable to spend those shares on (which would essentially act as a huge dividend at that time, however, most of the value of that dividend would have been already factored in the price of DNS since shareholders would be expecting that max supply reducing hard-fork well in advance).
Second, everything I said in the previous paragraph isn't how I would ideally
do it. I haven't been shy about discussing how I would ideally structure
. But the previous paragraph is IMHO a sane way of quickly restructuring the existing system if you are unwilling/unable to implement the proper foundation of delegate proposals ratified by shareholder vote.