Warning: this post is old and might not reflect the current state of the art
Latency Compensation is part of the Meteor Principles. There, it’s described in this way:
On the client, Meteor prefetches data and simulates models to make it look like server method calls return instantly.
What does it mean?
On a typical Web Application, when you do some kind of action, the action is passed to the server to be processed, and then you need to wait until the server responds, and then changes are applied to the page you’re interacting with.
More modern applications rely on AJAX to provide a better turnaround and avoid refreshing a page on every action, but still many apps rely on the server response prior to taking any action. Better apps introduce some sort of latency compensation, but it’s a manual process.
Meteor introduces the concept of Latency Compensation deep into his philosophy and it’s enabled by default, without you needing to do anything special to work with it.
For example when you add an item to a collection, meanwhile the item will be sent to the server, it’s already added to the collection view on your page.
It feels better, because the app feels immediately responsive (it is).
If there is an error you’ll be notified later and you have the opportunity to handle things in the best way for each situation.