come on ciyam.. you and me both know that we as individuals can read code.. and see if the code has anything dodgy in it (which some implementations have)
you and me both know that changing our current implementations to accept 2mb is only a few lines of code..
you and me both know that the hardware cost for 20 years of 2mb full blocks storage is only $100
so we both know that 2mb blocks are not the doomsday scenario. because we know blocks wont automatically be 100% full to 2mb from day one, but would take time over months for miners to adapt and test the waters..
so we know the only real debate is which dev team to trust to give us clean code.. or if their implementation of the bitcoin program also include other features that can cause abuse to fungibility or rarity, etc.
so how about open a github of an exact replica of bitcoin-core 0.11 with the only change being a GUI dialog box that allows users to set their own block limit..
imagine if everyone made a github replica that had just the feature of user controlled evolution.. and the community that can read code vetted it to say which versions were clean of any other dodgy features..
afterall if there are 10+ different implementations and 8 of them are clean.. then the 2 dirty devcrew would soon be weeded out and people would stop using them.. rather then debating which one of the 2 bandcamps to choose.. there would be 10+ different implementations with 8 being replica's of 0.11 with just the block increase tweak..
people wont need to debate which bandcamp of wacky /abusive features some deceptive corporate shill may also have hidden. as there are more then just 2 choices..
people wont need to debate control issues around there being only two working implementation and killing off innovation
people wont need to debate relying on that one working implementation being reliable to need to upgrade to later.. because they wont need upgrades, just setting tweaks..
so then people can change limits themselves when the needs must. without heartache, debate or constantly downloading new implementations in 2017 onwards.
in short.. having 10 differing implementations that all consensus to a 2mb limit is better then trying to get everyone into one of 2 bandcamps owned by corporations or have hidden agenda's
and later.. when the community as a whole have settled down and can all say they have the ability to increase to 2mb.. its a simple change of value in a dialog box.
and when the networking handshake shows that the majority of the community is 2mb ready because the majority of users have the setting set... then and only then will miners start churning out blocks bigger than 1mb but less than 2mb without problems.. because they wont want to do it before-hand as their blocks would get thrown out and their coinbase wins will be deemed unspendable.. so its not in their interest to churn out blocks that users wont keep
im not saying miners will fill blocks to 2mb instantly.. it will, in reality be something like 1.025mb at first which wont impact their hashing time.. and slowly increase as miners get comfortable adding more transactions to blocks (dipping their toe in the water).
edit:
there wont be multiple chains because miners wont screw themselves over mining blocks that the majority of users throw out. so miners will only do something different if there is consensus..
so there is no harm in having 20 different implementations as long as the main block checking rules all match and thus everyone is working off the same chain because their rules are compatible (consensus)