I saw this from Steve Barbour today. It is thought provoking and worth the read. I don't think I agree with him. But I like how he is thinking about the purpose of running a node.

I do not run a node because bitcoin is in a stable, secure economic environment.

99%+ of users do not need to run a bitcoin Core client and self-verify in such a secure environment. Do it if you like, but few will.

Indeed, 99% of bitcoiners today do not actually use a node ever. Thats because they haven't needed to. Relying on third parties (block explorers etc) to verify a transaction is just fine for most people, for most transaction values, for most of the time.

The main point for most users in being able to run Core is to be able to rapidly and reliablly sync to the chain tip to self-verify during insecure economic environment with high risk of 51% attacks.

In such an insecure economic environment users protect themselves against a doublespend by running a node and increasing the depth of confirmation before they self-verify accept bitcoin as a final payment.

Contrary to popular belief, miners do not secure users. We secure ourselves by choosing when to consider a bitcoin payment as final (confirmation depth user preference).

Contrary to popular belief, even in a 51% attack environment users can still use bitcoin, provided they choose a confirmation depth that outweighs the depth of a reorg. Yes, users have to take risk, nothing is risk free.

For example, if I sold my house today (secure environment) for bitcoin, the value is so immense to me that I would wait at least 24 hours worth of blocks before I signed over the deed. Maybe more blocks even. This is probably excessive, but I am paranoid.

If a 144 block (1-day) reorg happened by a 51% attack out of no where, I would probably require 2 weeks of block depth before I would accept the payment. Hard to say, but it would be deep.

51% attacks damage the utility of the chain, but they cannot kill the chain or network. Honest miners keep building on the chain tip, the chain must grow now matter what and cannot be halted.

A 51% attack would probably cause the price to plummet, but it would be still possible to conduct trade with bitcoin, just ultra painful. The price would stabilize vs the relative utility of the network, which may be temporarily reduced.

Running a node is most important when you actually need to run it. This is why optimizations (reduction in the sync time / compute resources required to sync to the chain tip) is extremely important. It is extremely important that the cost to run a node is minimized for everyone.

Bitcoin Core continues to optimze sync times, resource loads, etc with significant upgrades happening all the time.

I will run Bitcoin Core myself as a routine only when we enter an insecure economic environment. Until then I am quite content trusting mempool (dot) space, or other options with good reputations, to verify my transactions that are not excessively high value.

No, I also dont want "spam" bloating the resource cost and sync yimes, but nobody has a solution for it. Well, satoshi did - its called the block size limit. And it works incredibly well!

I strongly support the Libbitcoin project because it seems to me it is hyperfocused on an archtecture desiged to be useful in insecure environments.

I also support Bitcoin Core, but i believe competition and alternatives is important. Developers are, and will continue to be compromised by the State / Central Bank interests.

Bitcoin security is still widely misunderstood.