You are getting way ahead of yourself.  Lets just deal with this 2mb Bitcoin Classic Fork, get some experience and real-world data with a relatively minor and simple hardfork, and THEN we can move on to tackling even bigger scaling issues.

And it may very well be that Bitcoin NEVER reaches those dizzying heights.   But my guess is that as long as we can get through the current period of brief stagnation - and we can begin to move forward in slow and sure steps, carefully and learning as we go,  then as time goes by we will step with a clearer footing, more surefooted, and then at a quicker pace, a brisk walk, soon jogging, and maybe eventually at a runners pace.

But let us not get bogged down in contemplating steps that are not even in sight yet.  Slow and steady, ever forward.
inline signature removed

Bitcoin's ongoing tx per second hasn't reached to that level. We don't need that size yet. What we need to do is to increase it to 2MB, then 3MB if needed to contain current tx volume.

and what will happen in the next years? Every time we need a block size increase we have to go to the same hard fork drama or to have another ragequit Hearn story?

check2fire:

According to this chart, a block size between 500 MB and 1 GB would be sufficient to achieve 2000 transactions per second (which was only Visa's average throughput several years ago now).  We would have hit this point in the year 2032 following BIP101's trajectory and assuming a high rate of adoption.

can ever happen to have a system with so much high block size. What node will handle this and who will run that nodes? We will talk about 10-20-30 nodes or 100 and no more.

Franky, this is exactly what we already did with Bitcoin Unlimited.  There are already 120+ nodes running it.  Have you seen it yet?  

Our next release will also have a feature where the node advertises its block size limit in the user-agent string, so that sites like bitnodes can scan the network to help to identify where consensus may lie.  

You have to be sure that the drama is also "real", after the next hard fork, maybe users will have a more open mind, or more into the reality.

Also, after some next hard forks, maybe there will be a dynamic size solution.

Again, I think that the "common user", between the choice of making the transaction on OT/LN off-chian solutions or on-chain (at the same price, example: 0.0001 BTC), he will prefer the off-chain solution the majority of the times. (because faster)

My guess is that an intelligently automated scaling option will be forthcoming at some point.  But for now the task at hand is sufficient.

im using my own implementation, ive looked into bitcoin core, bitcoin classic, xt but not got round to BU.

to be honest it shouldnt matter what 'brand' it is as long as it does the job. so if your sourcecode is clean (as that the main debate everyone is having about all brands) then it solves some future debate issues by not having to keep downloading new versions for every fork debate

is really based in human desires or to mathematics? Satoshi really create a system to trust humans or to remove humans from the trust systems?