fix(server/impl): rate limit latent event acks separately - #4246

Acks for server-to-client latent events arrive through the same msgReassembledEventV2 handler as client-sent fragments, so they consume the latentEvent limiter (75/s, burst 125). The client acks every fragment it receives, so a transfer above ~150 KB/s exceeds that limit. The excess acks are silently dropped, and the server keeps resending fragments the client already has. Large latent events become much slower and waste bandwidth.

Fix: check the limiters after parsing, and give acks their own latentEventAck limiter (1000/s, burst 2000, configurable through rateLimiter_latentEventAck_rate / _burst). Acks over the limit are dropped silently, never kicked. Client-sent fragments keep the existing latentEvent / latentEventFlood limits unchanged. Malformed packets no longer consume any budget.

This PR applies to the following area(s)

Server

Successfully tested on

36897

Platforms: Windows

Checklist

• Code compiles and has been tested successfully.
• Code explains itself well and/or is documented.
• My commit message explains what the changes do and what they are for.
• No extra compilation warnings are added by these changes.

Fixes issues

#4244