fix(libp2p): gate connection-monitor abort on transport silence

[tabcat]

Description

Connection-monitor currently aborts on the first failed ping. With the AdaptiveTimeout 5s floor, transient loss or backpressure on a healthy connection routinely trips the abort even though TCP would have recovered (see #3463 for trans-Pacific evidence).

The default abortConnectionOnPingFailure: true is preserved but the abort is now gated on actual transport silence:

• @libp2p/interface: optional lastReadAt?: number added to MessageStreamTimeline; new ConnectionStaleError export.
• @libp2p/utils: AbstractMessageStream.onData() sets timeline.lastReadAt = Date.now() whenever bytes arrive — covers every transport in one line since they all route through this method.
• libp2p: connection-monitor adds connectionStaleTimeout (default 60s). On ping failure it aborts only if Date.now() - (conn.timeline.lastReadAt ?? conn.timeline.open) > connectionStaleTimeout. Aborts emit ConnectionStaleError so consumers can distinguish gate-driven aborts from transport-level failures.

Net effect: ping failures from heavy traffic on other streams, GC pauses, TCP retransmit bursts, or path-level jitter no longer kill healthy connections — only sustained silence does.

Notes & open questions

Some structural cleanups (drop AdaptiveTimeout from this file, simplify pingTimeout, drop abortConnectionOnPingFailure) become natural once the abort decision lives elsewhere — they're breaking and noted for v4.

Change checklist

• I have performed a self-review of my own code
• I have made corresponding changes to the documentation if necessary (this includes comments as well)
• I have added tests that prove my fix is effective or that my feature works