And it varies by year. There are years in which there is one consensus star (Bryce Harper, for example) and there are other years when any of 2-3 players could be picked first (Buxton vs. Correia).
Because basketball draftees play in the NBA immediately and are usually much more obvious, the first pick is much more valuable. There are probably data on the career WAR from #1 MLB picks vs. #2, #3, #4, etc. I'd bet that they aren't all that different, but that #1 is a lot higher than, say, #15. Tanking to get into the top 5 makes more sense than tanking to get #1 vs. #2.
If anyone has the data, please post!