Is there some technical reason I’m missing for why we want to shield the operators from any dilution?
At the end of the day, if we go this route, then we choose to generate capital from all holders by diluting the supply, basically levying a tax on all through inflation. Usually a characteristic of inflation is that it hits everybody equally, sharing the burden equally so to say. If we selectively exclude one demographic, then the other side of the coin is that we actively select another demographic to carry a larger burden. The kicker of course is that the token holders do not even get to have a say in this, so it will be the operators choosing to dilute the token holders. Not only will they get diluted, those saving up for nodes will be hit by a double whammy because now they will also need more tokens to run nodes in the future. This just leaves a bad taste for me, and I don’t really see the point except pure self preservation from operators, unless I’m missing something.
Also it just feels odd to now require an odd amount of tokens for a bond (e.g. 115,625 REN). How will this work with the burn mechanisms that have been suggested, do we decrease the bond again?
The 10,000 nodes was always just a magic number that felt like the right ballpark as explained by the founders. A much more elegant approach seems to be to keep the bond at 100k, do not compensate operators for dilution as this would dilute the supply even more, and simply allow there to be slightly more nodes in the future. Would it really be a problem if there are 10-15% more potential max nodes? Only about ~20% of possible nodes are ever online anyways, this means operator profits will already be diluted less than token value.
Last point, did we consider potential negative side effects of airdropping tokens to a subset of the holders? To me airdrops always seem to have negative externalities in that you get a tragedy of the commons type situation where it will be in your interest to dump the airdropped tokens on the market before the inflation is fully priced in. I suppose this is counteracted by the fact that there can be a lock-up period, and operators will need these tokens to keep running their nodes, but this too just feels like unnecessary complexity without much benefit to the project.