The language "save my info for later", and versions of that, are much better. It flips it around in the user's mind: "oh, now they're not asking me to give something up, they're helping me out"
While I understand what you're saying, I believe that would trip me up. I'm very much used to scanning for "Register", "Sign up", "Join" or "Create account".
I'm also unsure how much people really hate giving up their information. If anything, people are looking for places to hand over their information: Facebook, MySpace, LinkedIn, etc.
I think the problem is really that the more steps needed to convert your users, the more likely it is a part will give up before the end.
But we had exactly the problem you're describing: People are so used to scanning for "register"/"sign up" that some even assumed we were in invite-only beta.
We eventually changed that to a "Try it now -- no sign up required" button, which works better. The wording and UI is still tricky though: "Try it now", "trial account" etc. carry the connotation that it'll cost something later.
We've been meaning to A/B test a straightforward sign-up form for a while and measure effects on both conversion & retention, but unfortunately haven't gotten around to it yet.
The services that use the 'save my info for later' technique are most likely seen as easier because it is similar to how you use your local computer. You don't have an 'account' on your laptop. (Well, technically you do, but most people just believe that is just a password) You simply save something in your documents for later. By making the user believe that they aren't tied to anything, but just saving data for later, you can make them more likely to store their data on your servers.