It's "essentially worthless" unless you use it to draft Monte Morris, whereupon you have one of the best young point guards in the league. & if one could project players accurately, you would always prefer to draft those undrafted guys you mention than compete to sign them post-draft.
Still, the numbers you cite speak for themselves. At the same time, to me at least it looks like your gleaning from them is imperfect.
Darius Miller was picked by NO in 2012 & has been there ever since. He's logged @5200 minutes. He made $2.2m this year. He's no great shakes, but I'd say NO has gotten value out of the pick, wouldn't you?
Moreover, the important question is how good the players are: Isaiah Thomas? Patrick Mills? Ramon Sessions?
To be sure, late-picked guys more often fall victim to roster numbers than those picked earlier. But, I'd say the more important question is what kinds of players are available at that point in the draft? After all, R1 picks also move on during their rookie contracts. As I say, the question is what kind of player can you get late in R2? & equally important is how much that player costs you.
Consider the following list of 10 pretty bad players: Stromile Swift, Jay Williams, Darko Milicic, Michael Beasley, Hasheem Thabeet, Evan Turner, Derrick Williams, Michael Kidd-Gilchrist, Jabari Parker, Brandon Ingram.
Those are 10 of the 17 guys picked #2 in the draft from 2000-2016. More than 60% of the players taken second in the draft. &, because they were picked early, they cost more on their rookie contracts & often make a ton of $$ no matter how badly they play (case in point: Jabari Parker, who made $20m this year).
Consider as well the following. Could one expect, typically, to trade the #4 pick in the draft for the #11 pick & the #21 pick in the same draft? I'd say so.
Suppose we construct a thought experiment: in each draft from 2010-17 I have the #4 pick. For the pick I can have the best player taken from 4-10 in that draft, whoever we agree that player is. Alternatively, I can exchange that #4 pick for the #11 & #21 picks. With the #11 pick, I can take whoever was the best player taken from 11-20 in that draft, & with the #21 pick I can take whoever was the best player taken from 21-30, again in that same draft. & again anyone we agree is the best guy.
In how many years do you think it would be worth making the trade vs. how many years would it not? First off the top of your head, then after checking.
(Note that this is a mental model designed to illustrate how value changes down the draft numbers. I.e. to determine what's more valuable. Obviously it would not usually be practical to actually do this, as there wouldn't be a trade partner.)
I'll put my answer in a spoiler:
Now... what if it were the same except that for that #4 w/ the same restrictions you got the #11 with the same restrictions & the #31, which you could use to pick anyone taken in R2 or who went undrafted. A tougher test to be sure! Would it ever be sensible?
I'll put my answer in a spoiler:
