Bitcoin is highly divisible, but it makes no sense doing so!

TL;DR The high transaction fees of bitcoin will make owning small fractions of bitcoin unfeasible, even when using the Lighning Network

The latest Iced Coffee Hour Podcast got me thinking about the value of Bitcoin again. Listening to Michael Saylor for the first two hours kind of convinced me, that Bitcoin, or at least cryptocurrency in general, is a great alternative to our current financial system. I have heard about Saylor in the past, but just assumed that MicroStrategy is a smart investment fund. Now, knowing that it is a software company and he a sci-fi motivated engineer made me think more highly about his opinion. However, since he is heavily invested in Bitcoin, I find his reasoning too one-sided. When asked two times, he couldn't imagine any downside to BTC except for a technical flaw in the protocol.

First off, the reasons why Bitcoin is great and in Saylor's opinion, better than every other crypto. It is basically the only network that isn't "owned" by someone. There is no single entity (person, company, non-profit) that can assert their opinion or be forced to comply with regulations. No idea why this is important, since every node operator and miner decides what fork to support. This, alongside scarcity, is basically the only reason for him why Bitcoin is better. All other reasons, like self-custody and fungibility/portability, apply to every crypto. However, self-custody may become increasingly rare, because of the reasons I am describing it below.

Now, the issue that I am primarily concerned with! Saylor mentioned that Bitcoin is "infinitely divisible". That isn't really true, but that doesn't matter, it is highly divisible. If we assume that the BTC/USD price will grow to 1M+ in the future, as implied by his assumption of the cryptocurrency market cap growing from 2T to 100T (1:45:42) it would be more likely for someone owning a very small fraction of a single bitcoin. The issue is, that doing so makes no sense at all. For your transaction to be put into a block by a miner, you need to provide a sufficient fee. You can send a couple of Sats to a wallet but wouldn't be able to spend it! If you set your TX fee to 1 Sat, your transaction would never be picked up, since there will always be someone paying more. The queue of would-be transaction will never be this empty!

In the Podcast at 2:20:21 Michael Saylor mentioned that in the future, BTC transaction fees might be 300k, and he equates this with a 1M FIAT transaction cost for a 10M property sale. The issue with this and Bitcoin is, that the fees don't scale depending on the transacted amount. The size of a single transaction to send BTC from a single-signature wallet takes up the same space in a block, regardless of how many BTC are actually transferred.

The Lightning Network does not solve this, since there still has to be at least one transaction on the blockchain that enables the LN channel. In practice, this would mean that the default would become a LN-only wallet or a fully custodial wallet. However, this wouldn't really be Bitcoin but "LN Bitcoin". The core ethos of Bitcoin was and still is/should be a decentralized way to transact, without the possibility of being censored. The Lighning Network reduces that decentralization and, at least in my mind allows for cencorship. By how much I don't exactly know. If someone could explain this to me I would appreciate it greatly!

As far as I understand, there is not incentive for miners to support a soft fork increasing the block size, since this would decrease their earnings. We are at a point where the block reward becomes an ever smaller percentage of their total income. Therefore this issue will exist forever and will hold back Bitcoin development.