By definition, the RD applicants are not the same as the EA applicants.
In practice, the applicants who tend to have their s* together the most, and whose applications are in great condition at the beginning of autumn senior year, will generally tend to apply EA. They can expect a “bubble” or cluster of outstanding applications in the EA round. Which means that admissions can fairly easily pick out those exceptional applications and give them an early admission offer.
Applications that are not exceptional get deferred to the RD pool. These are often applicants who apply EA mistakenly thinking it is easier to get an admission offer, when their application really isn’t exceptional.
No, there is no reason to think that an application they are less than sure about should have a better chance than an application that hasn’t been seen before, because (again, by definition) you have no idea how good or bad the next application you look at for the first time will be. If an application wasn’t deemed good enough to merit an offer in EA, why would you expect it to be better than most of the RD applications? Submitting an application by an earlier date does not do anything to make the application itself any better or worse than submitting it later.
If UVA received fewer (or more) EA applications than expected, or if they found fewer (or more) exceptional applicants among them to admit EA than expected, they might need to see the rest of the RD applications before they can really be comfortable in putting together the incoming class they seek.
These are not spins of a roulette wheel. These are unique applications, each one representing a unique applicant. Ordering them would be a Sisyphean task is they tried to do it straight up. But they’re not even doing that; as you can see, each college (engineering, architecture, etc.) is looking for some target number of new students; someone trying to get into engineering is not compared directly to someone trying to get in to architecture. Or maybe the marching band needs 10 new trumpeters, and the orchestra already has too many violins; the next violinist might get rejected even if he has better grades and test scores than the next few trumpeter applicants.
If they’ve already blown past the (post-yield) quota of new engineering students, well, it’s full. If not, they’ll take the best applications for engineering they find among those remaining. And no, it won’t make any difference that they already looked at the application once a few months ago.