Thw problem with concrete limits on dilutions is that there already are: there's no dilution. Our ability to set social consensus disappears forever if we dilute btsx.
Well you can do anything you want with a DPOS network represented by a particular token as long as you get majority delegate approval. You can do anything you want with any blockchain system period through a fork if you are willing to consider the token of the forked system as "the same" as the pre-forked one, which is only a valid assumption if the market consensus decides to switch over nearly all the value from the pre-fork token to the post-fork token.
My argument is that we need to provide a mechanism to let the delegates know when it is okay (meaning too many shareholders won't all suddenly dump that token) to move over to some new version of code that hard forks and changes blockchain rules. Right now we have no blockchain-based mechanism. Dan tells the delegates to fork and they follow orders because they have no better option (that is currently the best mechanism for delegates to know whether shareholders will approve of that action because we are all going by the fairly correct assumption that BTSX shareholders trust Dan). If we want to eventually make this system more decentralized we need to provide an alternative mechanism (I think shareholder ratified delegate proposals are that system).
Thus, with such a system we do a one time switch of the social contract to: any change in the network is "acceptable" (acceptable to the shareholders, maybe not by you personally, if you disagree you can always sell) as long as it is either approved by the current rules of the DAC or by the majority of shares. The current rules (meta-rules if you will) of the DAC can create ways that allow for more agility in changing the rules of the DAC to meet the needs of fast changing scenarios without being weighed down by the burden of majority shareholder approval. For example my 15% * (100% - (percentage of shares voted)) net approval needed rule for changes to how the DAC revenue is distributed. Even rules whose validity is still depend on outside social consensus are possible, such as an elected group of N people which get to vote on hard forks (which are approved if at least M of those N vote to approve it) that don't change the human-understandable rules but rather just fix bugs in the implementation of those rules (if shareholders later find out the hard fork intentionally went beyond a bug fix, they can punish those M people by removing them from power and getting rid of their income). And of course, all of these meta-rules can always be changed if there is majority shareholder consensus.
Edit: If bytemaster really has a philisophical objection to my proposal system and not just a technical cost/benefit one, then I can actually to some degree understand what I think is his view. Providing a solution to quantitatively figure out the consensus of BTSX shareholders at this early stage might actually hurt the growth of BTSX. The lack of this capability means BTSX shareholders need to rely on the next best thing available currently: trust bytemaster unless there is VERY clear dissenting views by reputable forum members (in which case the delegates
might hold off on the hard fork). And let's face it, although making DAC decisions based on shareholder consensus is the most
fair way, it is not necessarily the
best way to grow BTSX value in these early stages (i.e. the benefits of democracy vs a benevolent dictator). That said, this capability eventually has to be built into the toolkit IMHO.