docs: correct stale doc-comments for LAN handshake and mesh-size

poll_lan_discovery's comment said Noise XX, but LAN-discovered peers dial over
UDP through initiate_connection, which uses Noise IK (IK at FMP). compute_mesh_size's
header comment still described the obsolete sum-of-disjoint-subtrees estimate;
the function OR-unions every connected peer's inbound filter plus self and
estimates cardinality once (matching the body comment). Comment-only, no
behavior change.

src/node/lifecycle.rs

@@ -901,9 +901,9 @@ impl Node {
         }
     }
 
-    /// Drain mDNS-discovered peers and initiate Noise XX handshakes.
+    /// Drain mDNS-discovered peers and initiate Noise IK handshakes.
     /// The handshake itself is the authentication — a spoofed mDNS advert
-    /// with someone else's npub fails the XX exchange and is dropped.
+    /// with someone else's npub fails the IK exchange and is dropped.
     pub(super) async fn poll_lan_discovery(&mut self) {
         let Some(runtime) = self.lan_discovery.clone() else {
             return;

src/node/mod.rs

@@ -1307,9 +1307,12 @@ impl Node {
 
     /// Compute and cache the estimated mesh size from bloom filters.
     ///
-    /// Uses the spanning tree partition: parent's filter covers nodes reachable
-    /// upward, children's filters cover disjoint subtrees downward. The sum
-    /// of estimated entry counts plus one (self) approximates total network size.
+    /// Builds an OR-union of self plus every connected peer's inbound filter
+    /// and estimates its cardinality once. Unioning (rather than summing
+    /// per-peer counts) deduplicates the overlap between the split-horizon
+    /// filters, each of which approximates "the whole mesh minus my subtree".
+    /// See the body for why all routing peers contribute, not just the tree
+    /// neighborhood.
     pub(crate) fn compute_mesh_size(&mut self) {
         let my_addr = *self.tree_state.my_node_addr();