Description
The cosmoshub document talks extensively about ICS1 (Simple Replicated Security) and how it can make the Cosmos Hub economic model (revolving around collecting fees from ICS hosted zones) work. While this is true, I actually argue this is not enough to make this economic model viable. In my opinion there is an underlying assumption - that demand for ICS hosting will be high, or that the chains that will do ICS hosting will have lots of transactions - that is actually arguable, and it somewhat contradicts the Cosmos thesis.
ICS1 is not VaaS (Validation as a Service) in my opinion, but a neat mechanism to scale functionality of the Hub. It is not an application in and of itself, it's just a scaling mechanism (for the Hub, not Cosmos). It's the gateway to having the cake and eating it too. You scale without losing security (so yes, I agree with the minimal Cosmos Hub thesis). As a ICS1 chain you are giving up sovereignty - or at least, you have autonomy but not independence - so it can't really be seen as a platform for hosting, as this is not really the internet (and for the hosting where you retain autonomy but not independence smart contracts are already perfect and will only get better). Migration in case the Hub decides to kill a zone and this zone wants to survive is not a smooth process, as it could be on the internet if a company has a problem with a cloud provider.
Or at least considering ICS1 as the way to generate profit in and of itself is not a realistic long term value prop in my opinion (but it's a form of usage of ICS1 that totally makes sense) for the Hub. Application layer is what wins, is where the value is captured. Applications will become so big to migrate to their own sovereign realm eventually. This is the Cosmos thesis, and the reason why Cosmos is the better model/design. It enables applications to become fully sovereign, which is what happened on the internet (big internet companies ending up creating their own infrastructure), and what will happen/is happening in blockchain. So again, I argue it is not sustainable in the long run. As a Hub functionality-scaling mechanism though, it's great and sustainable. ICS1 chains should be more seen as a part of the Hub, rather than a customer/third-party. But that means that the Hub itself should focus harder on the killer apps - and use ICS1 as an enabler. Of course they should be well designed and explored properly as if they were features to be implemented on the Hub itself (because in some sense they are).
For example solving the IBC routing problem is a huge component in my opinion in scaling Cosmos and also is potentially a place where I see the Cosmos Hub - using ICS1 - have a role. IBC needs to become profitable though somehow, which afaik it still really isn't. There is work but it's not here yet.
I agree another killer app(s) is smart contract platform(s). Probably desireable to have a consumer chain for each popular smart contract platform.
As a side note, something like Celestia might help the recovery of failed chains - or rather "DA hubs" could help. the Cosmos Hub is a security hub, and it does not really offer that service. At least not directly. What do you mean with "will be protected when it needs intervention"?