I'm as pro-analytics as anyone, but this article uses them as a be-all-end-all instead of a portion of the larger picture. While usage rate and assist% indicate the players who tend to initiate offense, they don't directly correlate to defining a player's abilities.
I'm firmly in the Adams starting over Gibbs camp because I think that's the best utilization of talent. And, this same data (along with the 'eye test') appears to support that. Gibbs primary skill on the court is his scoring ability. Throughout his career, he's been relied upon as the primary scorer for his teams. However, this is UConn, not Seton Hall. The talent level now surrounding Gibbs is superior, and he looks like he struggles to get into a rhythm when the ball isn't in his hands. On the second unit, Gibbs would function in a similar manner to his time at Seton Hall. He'd be relied upon as the primary scorer and facilitator. The ball would be in his hands and he could focus on finding his own shot; a role where he's excelled in the past. Even coming off the bench, Gibbs would still play close to starters minutes, as he'd be in our small-ball lineup.
On the other hand, Adams operates more like a 'true point guard' than Gibbs, and that's supported both by the eye test and the same usage rate and assist% data. Jalen's ability to get into the lane creates opportunities for his teammates, and his superior defense leads to more stops and fast breaks, i.e. easy buckets.
Adams and Gibbs can and will play a lot of minutes together, but I really believe Gibbs' talent is best utilized in a sixth-man scoring spark type of role.