When you get above T1 speeds, it’s no longer a “little” latency on two-way geo-synchronous satellite links. And T1 is barely considered broadband today. The faster you go, the worse it gets, unless it’s totally one-way. You have to rely heavily on WAN accelerators and hope your users aren’t doing encryption—because you can’t compress encryption. Wide area broadband should be done with stationary platforms (blimps) at 10-15 miles altitude or low-earth orbit satellites at 150-200 miles, but LEOs are expensive and complicated (Iridium).
It’s very important on long, fat pipes to adjust your TCP window size - this represents the amount of data that you can have “in flight” at any given time on, say, a file transfer or web browser connection. Most systems default to 32k, which on a high-speed 92,000 mile round trip means that drips and drops of data would be going across the link.
I would hope that the Wild Blue techs and docs would instruct people on how to fix this, but here’s a site:
http://www.psc.edu/networking/projects/tcptune/OStune/winxp/winxp_stepbystep.html
There’s a calculation called the “bandwidth-delay product” which can be used to determine a minimum TCP window size.