What do you mean by “capacity”? The capacity of the block? If we increase its size, the median fee will decrease. The miner will be able to accumulate more coins than before and the users will pay less, since their transaction will take less space.
That is actually quite a common misconception. Most people don't consider the elasticity of Bitcoin transactions when making statements. Bitcoin transactions are very elastic, if it doesn't prove to have far more benefits than other payment methods, they won't use it. This is the main issue that we're seeing right now. Price of Bitcoin is so high, yet actual transaction volume and real life usage is far lower than it should be. Fees has been high for a long time and it is obvious that the network won't be able to sustain larger TPS, and users have to resort to paying more fees.
The miners will someday, indeed, substitute the block rewards with the fees. That's obvious. The problem is that while their reward cuts in half every few years, they'll earn less coins and thus, they'll have less incentive to secure the network. It doesn't matter if a miner earns 1 BTC from fees plus 0.09765625 BTC as a reward. The fact that they'll earn less and less overtime means that the resources used in PoW will drop off. And that's true if we assume that they'll fill each block with transactions while the price, of course, remains stable.
4 years is a long time. There is a fixed equilibrium where miners starts to drop off as soon as it becomes less profitable for them. There is a reason why the decrease in block rewards is the most significant at the start and starts to taper off as we reach more milestones.
For instance, what will happen if there aren't enough transactions in the mempool? If the miners' income depends mostly on the fees, then there're lots of factors that affect the practicality of Bitcoin.
In 8 years, if we're facing a problem where we still don't have enough on-chain transactions or still cannot meet the needs, then I suggest that we should just ditch Bitcoin as an experiment.
I just want you to confirm me that the incentive will decrease over time and therefore, so will the security.
I don't agree that the incentives will decrease to the point that it becomes a threat to Bitcoin's security. The block rewards will decrease over-time, that isn't really disputable. Just look at the numbers. I don't agree that the drop in block rewards cannot be cushioned through other means.
I'd just like to seek your understanding on this concept; There is no reason for concern until we reach a point where it becomes easy or feasible for an attacker to attack Bitcoin. Do you agree?