What verification actually prevents

It is a fair question. The answer is narrow, and being clear about how narrow makes the habit easier to keep rather than harder.

The specific failure it removes

Accepting an address from a source that is not the operator. That is the entire scope. It does not make a marketplace trustworthy, does not protect a payment and does not predict whether a platform will exist next month.

Why that narrow scope is worth the effort

Because the failure it removes is the one that actually happens. People do not usually lose accounts to sophisticated attacks. They lose them by arriving at a convincing copy during an outage, and every defence except this one operates after that has already happened.

How it compares to checking the address bar

Comparing the onion on a login screen against the bar catches a copy you have already opened. Verifying a signed announcement catches a bad address before you open anything. They protect at different moments and the second is strictly earlier, which is why it is worth having both rather than choosing.

When saved addresses are enough

When nothing has changed. An address you saved months ago and have used since carries continuity, which is weaker than a signature and perfectly adequate for routine use. Verification earns its cost at the moments of change: a rotation, an outage, a first visit, or any address you did not have yesterday.

The honest cost

Ten minutes of setup once and about thirty seconds per announcement afterwards. Anybody describing it as effortless is overselling, and anybody describing it as impractical has not done the setup.

What to do if you will not do it

Then keep a saved list religiously and never take an address from anywhere else, particularly during an outage. That is a weaker position honestly occupied, which is better than a stronger one performed badly.