Sustainability can be simply defined as the protocol staying online, resilient to attacks, and usable under all conditions. It also needs to be relevant and keep up with contemporary needs, so to speak.
Different types of projects have different requirements to achieve this goal.
Of course, the most challenging is L1 with a proof-of-stake/proof-of-work consensus mechanism.
L1
Economic aspect: a) subsidy > operating cost of the entire network; b) demand for basic assets (demand drivers can be varied) ≥ subsidy
Technical aspects: operating costs increase linearly; verification costs remain almost constant
Social aspect: multiple development teams, a large number of organized individuals who can implement changes
Clearly, social sustainability is paramount. A strong social layer can overcome the challenges of technological and economic unsustainability. But without adequate social participation, a temporarily strong economic and technological project can fail as conditions change.
In addition to sustainability, the network can also thrive if demand exceeds supply, which leads to greater economic security. Alternatively, verification costs decrease over time. Or, it becomes so incredibly socially diverse that leaving a large part of the developer ecosystem doesn't make a noticeable difference. However, minimum sustainability should be a basic requirement.
If L1 can achieve sustainability in all three domains, in all possible scenarios, that opens the door for different types of protocols. We call this sustainable escape velocity.
You can have an immutable smart contract infrastructure like WETH, which fully inherits the sustainability of Ethereum.
The next layer up are immutable smart contracts like Uniswap, but they do face two additional challenges:
a) The front end needs to be sustainable and have a variety of front ends
b) Ongoing (technical and commercial) development is required to remain relevant
Both a) and b) are possible through tokens, although UNI is currently clearly a poor example of this, with no clear sustainable demand driver. This is also where you need social sustainability - a motivated community working to make the UNI token sustainable.
Of course, other projects like Aave, Lido, Maker, ENS, etc. do better. Maybe they need to because they are not immutable and now you need your tokens to be secure. To be clear, their need for economic security is not as urgent as L1, because there are additional checks and balances, such as time locks, strict approval/voting thresholds, etc.
There are various dapps and they will have different demand drivers, different operating expenses and subsidies. Either way, the key is that they need to ensure that demand matches supply or exceeds supply.
By the way, when I talk about sustainability here, again, crypto protocols should be encouraged to be profitable — not just to survive, but to thrive. There is an evil “crab in the bucket” mentality (“if I don’t have it, you can’t have it”) in the crypto space, and profitable protocols are frowned upon. Without antitrust regulations and strong economies of scale, especially for financial dapps, the crypto space has always tended to form oligopolies — this is just a side effect of the openness of the crypto space. Ideally, we should have layers of society strong enough to deal with social pressure from oligarchs (actually implicit antitrust measures), although this may not be realistic.
L2
An immutable L2 requires almost nothing more than an orderer and an honest validator to survive. But in reality, most L2s won't be immutable, some will have multiple orderers, and will spend a lot of money subsidizing protocol growth (like Optimism's public goods funding). Fortunately, L2 can experiment with different revenue models, especially application-specific L2/L3 etc. We’ve seen some interesting examples like Immutable X or dYdX, which have no gas fees but monetize through transaction fees; or Sorare, which has a business model that’s directly application-specific. For general purpose L2, they certainly need to consider user transaction fees and MEV, but also novel models.
Regardless, the same rule applies to L2: the higher the income, the more they can grow and thrive.
All in all, crypto projects end up not breaking the ground rules. Just like enterprises and countries hundreds or thousands of years ago, there is no such thing as creating value out of thin air. Value must be obtained in different ways that are appropriate to the protocol.