Agreed that both stats are interesting. However, the tweet 'implies' that Undrafted rookies have a good chance or that there is hope for them to become starters, "Hope for underdogs:..." based on the total number of Undrafted Rookies compared to 1st and 2nd rounders. In actuality, 1st and 2nd rounders have a much better chance of becoming starters. Its a classic case of (mis)using statistics to make a point.
I re-read Schefter's tweet and there's no mention anywhere of starters, just roster spots, so I think that implication wasn't Schefter's.
I certainly do agree that 1st/2nd rounders have a much better chance of becoming a starter, but I'd also argue that finding an UDFA as a starter has a considerable amount of value, perhaps just as much as hitting on a high draft choice, because of the cap savings.
Consider a 1st-round CB like Artie Burns working out. It costs you a high draft pick and if it works out, you're paying over $9M for 4 years, over $7M guaranteed. A 2nd-round pick like Cyrus Jones is considerably less, say 4 years for $4M ($1.5M guaranteed). But an UDFA like Malcolm Butler is 3 years, $1.5M with no guarantees. So there's less risk, less guaranteed money, and less overall money. It's also harder to find a guy who can start, but when you do, it's an incredibly valuable find and extremely helpful for your team's cap situation.
And sure, it's not easy to find UDFA starters, and you can't expect to do it consistently every year, but there's considerable value when you do.