Followup:
Did some diagnostics on my connectivity.
Pinging Google:
Code:
Ping statistics for 146.115.8.113:
Packets: Sent = 2000, Received = 1999, Lost = 1 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 4ms, Maximum = 42ms, Average = 9ms
Pinging my world server:
Code:
Ping statistics for 199.91.189.44:
Packets: Sent = 2000, Received = 1932, Lost = 68 (3% loss),
Approximate round trip times in milli-seconds:
Minimum = 38ms, Maximum = 113ms, Average = 45ms
Packet loss appears to be an issue. I opened up wireshark to try and capture what was going on. Here's a quick graph of a "no-lag" section:

And one where there was sufficient lag to produce a 90k error:

The red impulses are TCP errors - namely, TCP Retransmission, TCP Dup ACK, and TCP Previous segment not captured. So this is indeed a connectivity issue - just one out of my immediate control. I have the wireshark log saved if anyone is interested.
I have enough knowledge to get this far but past this I run out of resources. Pinging google is sufficient to show that my immediate ethernet is not malfunctioning, but there are a lot of stops in between me and the world server, and any one of them - including the world server - could be dropping packets.
I will keep this thread updated after the server maintainance to let interested parties know if that helps the connection quality. If it does, great, thread closed; if not, I'll start combing through the traceroute to my world server to try and pin down who is dropping packets.