Yes, I mentioned it in other comments that it could also end up being a layer 2 or sidechain for Tron. What we do know is they will integrate Steem with Tron in some form or another, they aren't going to keep developing it in parallel.
The reason why it makes sense to build one network is because all resources can be combined into scaling and developing one protocol. Note: One protocol obviously doesn't mean just one chain, e.g. Ethereum 2.0 is going to start off with 64 shards (basically, sub-chains) that execute in parallel, with a view to increasing it to 1024 in the future. One protocol is also a combination of layer 2 solutions, sidechains etc. See what I said about network effect and economies of scale. Tron don't want to waste time building a second protocol, they just want to keep developing Tron as the one platform to do it all. Of course, this can take many forms, like you point out.
There are certain exceptions with very specific case blockchains, but even then, it makes a lot of sense to be layer 2 and sidechain, and leverage the network effects of a consolidated platform. By the way, Steem is not it, there's nothing special about this chain - it works like any DPoS chain. The only difference probably is that it stores a lot more text than transactions / contract executions than your average blockchain, but technically that's a non issue.
I'd definitely have to disagree on that one. The tech behind it is actually really genius. Lots of graphene innovation has come from Steem, lots of additional features to match the ever online social media world, and overall it has a lot of concepts adopted into other places.
Looking towards Ethereum for comparison, their sharding methodology is vastly underperforming. They predicted what? 5x less shards with 10x more output? But that was no where near the case. Ethereum has a loooong ways to go tech wise, EOS has a loooong ways to go decentralization wise ( and currently usability wise ), etc.
Tron, and Steem, both have problems. But from what I've researched we are positioned to iterate faster than the competition and even in our less-than-perfect state we're still outperforming half the other chains out there ( more than half, but I don't wanna suck too much Steem dick tonight ;^) ).
Steem may not even be a "sidechain" as we know them today, perhaps it'll be more of a "complementary chain". No "one chain to rule them all" ideology, maybe the best bet is to allow multiple networks catering to different areas. Steem would be virtually untouched in this case.
Yeah, we'll have to agree to disagree. EOS 2.0 is vastly superior to Steem at this time. Which should surprise no one, because Steem has hardly seen any development in the last 2-3 years, whereas EOS has taken the Steem codebase from early 2017 and poured in probably 100x development resources to improving it. Meanwhile, Steemit has spent years chasing a scalability rabbit hole which still remains unsolved with the broken/limited MIRA implementation which no one uses. I rejected HF20/21 for a variety of reasons, but one of the reasons was the pathetic state of affairs with Steem development. Mind you, all of this is fixable. I don't know enough about Tron yet to comment. In general, as I mentioned before, all of these DPoS chains - EOS, Steem, Tron - are fundamentally flawed and unsustainable.
As mentioned in my post, Ethereum's Layer 2 solutions are already excellent. Optimistic Rollup is capable of hitting 2,000+ TPS, which firmly puts it in EOS/Tron territory, all of it without cluttering Layer 1 with unnecessary data, and of course, the DPoS challenges. Loopring powered WeDEX is even better than NewDex on EOS, and light years ahead of Steem Engine or the internal Steemit market. You even have DPoS-based layer 2 chains like Loom. Ethereum 2.0 is well on track to offer all the scalability Layer 1 will need for the near future without any of the fundamental compromises of DPoS, but of course, it's still a couple of years away. Same applies for Caradano, but I fear they're too late.
But the secret, of course, is that scalability is a red herring. What makes a blockchain successful economically is high value transactions. High volume, low value transactions are ill-suited for Layer 1 blockchains in the first place.
You're probably going to disagree with everything I've said, and that's fine, let's move on.
This