There are a lot of variables at play such as ballpark, roster context, team strategy, ballpark, and AL/NL, so there's not necessarily a right answer. Your primary goal is for your leadoff to score a lot of runs, so guys who get singles/walks then can either go 1st to 3rd on a single and home on a sac fly, or steal second and score on a single, those are obviously useful things. BR IQ + speed, contact + splits, eye are all key attributes. Imo IQ is better than speed
Player Profile: Christian Ross on my AL SF team is your more traditional high-speed, low-ish power, .285/.350/.390 role guy. Ross is kind of platoon-ish, but he does a decent job bunting his way on base against LHP so it offsets itself a bit ... doubles seem very low for his vR but triples are good because AT&T is +3 for the asymmetric geometry. SB volume is excellent but SB % is not great, getting a little unlucky there maybe. That team is stacked with speed guys throughout the lineup so I go 5-stealing 5-aggressiveness and lead the league. I would expect better run volume from him, never had a 100-run season, but he is limited by low DUR and platooning. A little disappointing that he isn't manufacturing more effectively but oh well. [As for IQ > pure speed,
Player Profile: Omar Flores is 58 speed and 99 BR and is super efficient, career 194 SB to 21 CS, but I do not use him as leadoff because his OBP .323 is too low for my taste. His ratio of OBP to runs is strong, he does have a 100-run season, and he's pretty good as a #2 or a #6/#7 hitter (AL)]
On my best NL team I prefer using
Player Profile: Nick Holmes as leadoff, he is a much more well-rounded, complete 5-tool hitter. The speed is there- he can swipe 40 bags, the power is there- he has 70+ extra base hits two seasons in a row, the IQ is there- he has scored 110 runs five times. At different points I have tried him at #2, #3, #4 and #5, but he simply is not as good in those roles as he is at leadoff. He actually kind of sucks in those other roles, even at #3 where he should theoretically be great.
In the NL I am more likely to play a traditional #3 hitter at #1 because the team is only getting production from 8 spots instead of 9, so it's a bit more important for me to accelerate my lineup and get runs right away. In terms of roster context,
Player Profile: Nick Terry was an excellent complement in the #2 spot, .297/.357/.426 was a bit of a dream for his skillset, he's a switch-hitting singles guy who chips in his share of doubles, so he was persistently moving Holmes from 1st to 3rd and from 2nd to home plenty of times on singles, and 1st to home on doubles, and getting on base himself. In the #3 spot, teams were simply refusing to pitch to
Player Profile: Jesus Espinosa in any context- .448 OBP is just ridiculous, but he was punishing teams when they did with .335 avg and .543 slugging. When your #1 and #2 guys are already on base, a walk for the #3 is worth up to 3 total bases instead of just 1, and when a guy is standing on 3rd a flyout is an RBI. Simmy actually nominated him as its primary choice for MVP even though he only had 99 RBI on his own. Behind those three guys were
Player Profile: Hughie Jakubauskas putting the ball in play and moving everybody around (80 XBH and 128 RBI), and
Player Profile: Michael Ni smashing home whoever was left over (37 HR and 110 RBI).
None of those players are "perfect" by any means (Holmes is by far the most well-rounded of them all but still not "perfect"), but we were able to use Holmes in a way that made the rest of our context more successful. I was definitely sacrificing Holmes' individual RBI potential, but in exchange for making Terry wayyy more useful, enhancing the threat of Espinosa, and allowing Hughie and Ni to thrive by just swinging away with tons of guys on base