When we talk about risk versus reward in the context of Celtic performance, we are looking at what you get on the upside in terms of goal threat and creativity versus the downside of lost ball possession or defensive risk.
One view to illustrate maps the threat posed in the final third in terms of expected assists plus expected goals (called expected scoring contribution) compared to the number of times a player gives away the ball in the final third without any advantage (e.g. winning a corner, taking a shot, setting up a chance, or even maintaining possession).
2022-23
As a starting point here is that view for last season.

The players arrange by positional population with a couple of highlighted outliers.
At the bottom are the full backs. All created roughly the same amount of attacking threat but with variable risk associated. Alex Bernabei and Tony Ralston tended to lose the ball with no gain more often that the other three full backs.
Moving up to the central midfielders. David Turnbull offered the highest open play threat. (By the way to remove the impact of set pieces from this analysis xSC covers open play assisting passes and non penalty xG only). Matt O’Riley had the highest possession loss and Reo Hatate the lowest xSC. Note Aaron Mooy’s contribution.
Top left are the wingers. Jota was the epitome of high risk / high reward wing play with the most final third losses in the squad but a high xSC. Arguably the most efficient was Liel Abada whose xSC is more akin to the strikers in the top right corner.
Finally, the outliers. Callum McGregor has the lowest xSC and lowest final third loss numbers. His role is specific to anchoring the midfield and even more so under Ange Postecoglou.
Lastly Daizen Maeda has a relatively low xSC more akin to a central midfielder and lowest xSC of that group. He is low risk / low reward as a winger but obviously brings a lot of off ball attributes.
Note overall the positive contributions by this view of Jota, Abada, Haksabanovic, Turnbull, Mooy and Giakoumakis. Why mention them?
Because they are either no longer here or long term injured or not part of the first team.
2023-24
Very early days this season and data sample sizes quite small.

Firstly, note that the axis boundaries are less favourable. Final third losses now go up to 12 per 90 minutes and Open Play xSC only goes up to 0.8.
We have broadly the same groupings and on the plus side Alistair Johnston seems to be relishing the role of a more straightforward marauding full back as opposed to the tricky inverted version.
Kyogo is in his own little island but xSC is down from 0.8 to nearer 0.7 and there is only him!
The midfielders are posting similar numbers but out of favour Turnbull has the best return for the least loss.
Of the wingers, Maeda continues to profile more like a midfielder whilst Abada is a long-term injury.
New players Yang and Palma are yet to settle in but are stretching the boundary of final 3rd losses whilst achieving xSC contributions more akin to the full backs.
Needless to say neither Jota or Mooy has really been effectively replaced as yet.
Summary
So, not to set alarms off as such, but until new players settle in and injuries clear, we are looking into quite a deficit of creative reward versus final 3rd loss risk so far.
Kyogo’s contributions are down, and he needs more support – Giakoumakis was very effective off the bench. Indeed, as note above, it was the five subs including Haksabanovic, Turnbull, Abada and Giakoumakis notably that often came on late in games and padded out scorelines and brought game to a close.
Celtic seem to have less depth and even last years stellar performers are yet to hit the heights of last season.
As mentioned, injury is a factor as is a less than favourable initial fixture cycle. New players may yet shine.
But Brendan Rodgers could be entitled to think he needed more depth to the attacking third threat.