[CTcard]

I can understand the argument for OL,  TE, and for sure RB. But QB? KJ has played part of one season. Let's not make the same mistake that was made with Chryst, in thinking a few good games means you are great.

I am not exactly sure who the we is that you are referring to, but for the most part the Stanford fan base was highly skeptical of Chryst going into the 2017 season. This was mostly because his three good games came against atrocious defenses, but also because he had collapsed into inaccuracy problems at several points in 2016. I myself argued on this board, that Chryst's good games were still good games even if the defenses were bad - but was also wary about what he had established and his consistency.

Costello does look a lot better than Chryst to me, largely for the reasons listed above but perhaps even moreso because he doesn't seem to go into a funk the way Chryst did from time to time.
If you want a statistic on the point, I'd go with:
In 2016 Chryst played the predominant part of five games, with two very poor performances. Four of the five were against very poor defenses, with OSU being the best at #79.
In 2017 Costello played the predominant part of eight games with one very poor performance. The poor performance was against a very good defense (WSU - #16 overall, #7 pass defense PER). Five of the defenses he faced were better than the best defense Chryst saw in 2016, and all were better than the second best defense Chryst saw.

-----
To the broader point, I would agree that one cannot declare Costello better without examining the other guy. 
San Diego hasn't thrown the ball much under Chapman, averaging under 19 attempts per game for three straight years. For comparison, last year Stanford threw 25.9 times per game.
Chapman has been pretty efficient - roughly what Shaw wants from a QB. Last year Chapman had an efficiency rating of 139.2 just about dead even with Costello at 139.5. The previous year Chapman came in at 149.2. 
An interesting (infuriating?) bit about Chapman was that his game versus Stanford was his career high in attempts and completions, his only game with over 20 completions ... only game over 16 completions for that matter.

[CTcard]

I can understand the argument for OL,  TE, and for sure RB. But QB? KJ has played part of one season. Let's not make the same mistake that was made with Chryst, in thinking a few good games means you are great.

I am not exactly sure who the we is that you are referring to, but for the most part the Stanford fan base was highly skeptical of Chryst going into the 2017 season. This was mostly because his three good games came against atrocious defenses, but also because he had collapsed into inaccuracy problems at several points in 2016. I myself argued on this board, that Chryst's good games were still good games even if the defenses were bad - but was also wary about what he had established and his consistency.

Costello does look a lot better than Chryst to me, largely for the reasons listed above but perhaps even moreso because he doesn't seem to go into a funk the way Chryst did from time to time.
If you want a statistic on the point, I'd go with:
In 2016 Chryst played the predominant part of five games, with two very poor performances. Four of the five were against very poor defenses, with OSU being the best at #79.
In 2017 Costello played the predominant part of eight games with one very poor performance. The poor performance was against a very good defense (WSU - #16 overall, #7 pass defense PER). Five of the defenses he faced were better than the best defense Chryst saw in 2016, and all were better than the second best defense Chryst saw.

-----
To the broader point, I would agree that one cannot declare Costello better without examining the other guy. 
San Diego hasn't thrown the ball much under Chapman, averaging under 19 attempts per game for three straight years. For comparison, last year Stanford threw 25.9 times per game.
Chapman has been pretty efficient - roughly what Shaw wants from a QB. Last year Chapman had an efficiency rating of 139.2 just about dead even with Costello at 139.5. The previous year Chapman came in at 149.2. 
An interesting (infuriating?) bit about Chapman was that his game versus Stanford was his career high in attempts and completions, his only game with over 20 completions ... only game over 16 completions for that matter.