This is not binary like this either. There are a TON of variables.

• You can have the IPs you communicate with visible to your ISP directly, or hidden from an ISP but visible to a VPN, or hidden from ISP but visible to the Tor network, the safety of which depends on “against whom”.
• You can have your messages encrypted in transit but visible to the messaging server, or encrypted end-to-end and thus useless to the messaging server too.
• You can have the identity you post under bound to an identity outright, or you could obfuscate that.
• You can use a centralized messenger that has your whole communication graph and all metadata, or you can use a federated one with multiple identities and thus metadata scattered across multiple places. Or Briar that doesn’t have servers at all.

All depends on whom you want to be private against, as well as how much effort they want to put into getting your information. There is no “absolute privacy”… But there is “requiring more effort from the chosen adversary than you’re worth”.

You’re either connected or disconnected. There is no in between. All you can do is toggle between them and hope no one is paying attention.

No I’m not. Google up police cracking criminal crypto wallets. These kinds of responses are exactly why this question got asked.

Do you need links to police cracking people crypto wallets. That’s about as secure as you’re going to get now and it’s still not enough. So what else have you got.