Why not adjust the block size depending on the load? Bitcoin is becoming more popular and this will inevitably lead to an increase in the number of transactions. A mechanism for increasing the speed of their confirmation is simply necessary. Without this, bitcoin does not have a future. If nothing is done then we can face the fact that bitcoin will give way to someone else. For example bch. I don't want to.
Because miners will rip all the profits while nodes will bear the burden, and this can be used as another vector of attack - by spamming the network with expensive fees, like its happening now, attackers will be able to weaker kick nodes out of the network, making it less decentralized. Bitcoin was created to make it impossible to take down, and with 80% of mining power being already controlled by a single entity, our network of nodes is the last line of defense against attempts to kill or takeover Bitcoin, and I'd say survival is more important than someone's $100 transactions that are most likely sent to an exchange to dump Bitcoin at ATH.
OK you are advocating survival but it is not constructive: you are saying don't do this and don't do that but you are actually not proposing any solution. I run a bitcoin full node AND at the same time I sometimes have to do a transaction so I think I am entitled to tell you that the disk size is not a problem, while the tx price is a problem and will kill bitcoin is nothing is done - that I can tell you for sure. And I share Lieldoryn's sentiment: I don't want to.
You are saying Bitcoin was created to make it impossible to take down. Well, it doesn't look that way to me
EDIT: if you adjust block size to the number of transactions, miners will NOT rip the profit. It will still be profitable to include every, even a small fee tx compared to NOT including it in a newly mined block. They may reasonably choose not to include only absolutely no fee transactions.
How can fees kill Bitcoin? They are only as high as people are willing to pay, it's an open and dynamic bidding on blockspace - if people would start leaving Bitcoin because they don't want to pay high fees, fees will go down because there will be less competition for blockspace.
Disk space is only a part of full node resource requirements, it also consumes CPU processing power, RAM, bandwidth, disk reading and writing. If you will increase the blocksize, all of those factors will increase in O(n), weaker nodes will be knocked off the network, more powerful nodes will notice that Bitcoin takes so many resources that it's impossible to run it in the background while they run their programs, and people who buy powerful computers are doing this for a reason - either work or entertainment, so they will be turning off their node software, which will result in lower uptimes. And all this negative stuff won't solve the scaling, it will merely kick the can down the road and fees will be high again in short time.
