Thanks for the thoughtful responses, José!
Tranched Model
I’m curious to read about the tranched model you mention. When we look back at the initial issues you sought to address – reducing contagion risk, expanding risk/reward diversity, boosting innovation – the tranched model does appear to move in that direction, but to a much, much lesser extent. Essentially it would offer 1 or 2 more levels of risk exposure, where the aDAO model could literally provide infinite. Would be curious to hear what feedback led to reconsidering the tranched model. I can see that from an implementation perspective it could offer a more manageable increment in complexity, but I sense you’re not shy about making larger changes to the protocol at such a nascent stage.
External aDAO Control
Funnily enough, this was precisely my concern. If Aave achieves protocol status, then it becomes a massive ecosystem player, whose decisions have major implications on the projects who implement it. Therefore, if external asset holders don’t have a vote on certain parameters, they have to bear the risk of a potentially arbitrary set of actors who don’t hold any skin in their game.
So I remain convinced that for Aave to achieve Protocol status, and the scale which is possible with that status, some element of ownership over aDAOs by the external assets they serve is essential. The simple reason is that Aave and credit in general is such an important component of the ecosystem and to the projects it integrates with, that an aDAO can become a risk to the project in and of itself. Take @Jordan’s recent post about governance attacks. Governance attacks don’t directly hamper Aave – they adversely affect the projects that implement Aave. You mention skin in the game, but the reality is that having skin in the Aave game is only one part of the problem – for implementing projects to trust aDAOs they need to feel like they have some control, too.
You mention the risk of using external assets to govern risk parameters that have potential to adversely affect Aave itself – to be clear, I’m not prescribing that. Concretely, the kind of parameters I suggest external asset holders can vote on are those that directly affect risk or functionality of external projects. For example, I suggested a binary option to allow external asset holders to vote on whether their pool issued a ‘marked’ token – a bToken – which could prevent governance attacks. Whether or not that idea has legs aside, it’s the kind of optionality which would inarguably make Aave a more desirable option for go-to credit protocol.
I recognise the idea of allocating aDAO tokens to external holders, but I must admit that feels like heavier work than simply using the assets themselves as aDAO governance tokens. Rather I would propose simply making it a parameter, votable on the aDAO – “for parameters X, Y and Z votes will be considered in tokens A, B and C.”
Determinism/Autonomy
Your answers to a couple of my questions around deterministic pricing and permissionless management for aDAOs were essentially “let’s leave it up to subjective consensus”, and I think that does feel sensible. Might be possible to push for greater determinism and autonomy later. 
Upgrades
A separate question which occurred to me – apologies if I’ve missed this in the text – is around what global upgrades would look like. For example, Ecosystem Governance decides that a new parameter is desirable for every aDAO template. How much choice do you imagine aDAOs having in deciding whether or not to upgrade? I’m imagining that if the ecosystem reserve and aDAOs are adequately aligned then the game theory should take of itself over time, but I wondered if you’d had any thoughts.
Enjoyed the video! FWIW I didn’t mean “dense” negatively when talking about the proposal 
Just meant very detailed lol.
Cheers!