paying higher fees is not the solution, it is like an escape temporarily for users to compete with each others and the spam attacks to get a faster confirmation. and things may not be as bad as it seems these days but soon there will come a time that real transactions will be stuck waiting for confirmation no matter how much higher fees they pay because simply there is not enough room in a block to include all of them.
It's a point I've raised many times before, but it seems to fall on deaf ears. People start talking about how it'll supposedly be better when the microtransactions are gone and it'll be just "legitimate" transactions (as if legitimate is something consensus has arbitrarily defined) remaining. But it doesn't change the fact that finite space is finite and full means full. In reality, in this economy, microtransactions should be seen as a "buffer" or "padding", so once they're gone, it means everyone queues when they get caught out by the guess-the-fee lottery. There's a sweet spot we ideally need to hit, where there aren't too many microtransactions and enough so that larger transactions aren't delayed. But some just aren't interested because they believe not having a buffer is somehow healthier. Plus no one really knows how to maintain said sweet spot, since network traffic is almost impossible to predict. Adaptive blocksizes are problematic to implement and it doesn't feel like progress is being made in that area.
OK, I just want to say something about this if you let me, about "legitimate" transactions, in the world in which we move some things are based on precisely the fact that need be defined but not always from the beginning, there are certain criteria that need to be fulfilled to make use of some thing or service, an example could will be more clear, in most of the countries vehicles are used where the steering wheel goes of the left side taking as reference to the driver and in others the right side, beyond which at some point there was no consensus to make all the steering wheels of the car on the same side, each one chose to make his own method, and that is what is actually being seen, solutions to scalability are already given by those who have focused on carrying them out, everyone can easily realize this, only need to check the number of nodes daily for a period of time and see which client are running the most of nodes, and the rest will depend on the performance of each network, although it was not related to scalability, ETH did this, if at any point it fails, we must not forget that it was mentioned at the beginning as an experiment.
