Let's say that the Celtics have replaced most of Kyrie Irving's production with Kemba Walker. I would argue that is a bit of a stretch given how efficient Irving is versus Walker - especially from deep which many NBA pundits will tell you creates gravity that helps teammates get better looks. However let's call that a wash.
The team still has to replace scoring by Horford, Morris and Rozier. While Morris tailed off horribly toward the end of the season and Rozier was inefficient, at best, both guys scored a bunch of buckets for the Celtics and were each able to, at times, create their own shots.
Its entirely possible that Hayward, Tatum and Brown each improve enough to up their output while Kanter largely replaces Horford's buckets and some of the kids step up too. I suspect we will see some of these outcomes too - it just seems overly optimistic to use when trying to make a reasonable prediction. I would also point out that while Kanter might be able to equal Horford's scoring, his lack of a three point shot and his inferior passing skills will make it tougher on his teammates for obvious reasons.
That said, if I am wrong and things work out better than expected with Arson Edwards winning ROY while TimeLord becomes the next great two way big en route to a 50+ win season, there will be no tears here.
This is a good discussion, so let me make it clear where I'm coming from: I do
not expect the Celtics to be better with the losses they've had--I think the 44-46 win type predictions are about right. My main difference of opinion is that I think the slide will happen more on defense than offense.
The reason I mentioned Charlotte's offensive results with Kemba is that they're a good example of how NBA offenses don't really consist of discrete components: they're ecosystems. Kemba for whatever reason was able to take a team with far less talent than Boston and get it to play almost as well offensively.
Relative to Kyrie, a lot of the things he does well are helpful for teams where the other guys aren't primary scorers: he keeps the ball middle more than Kyrie, he's more active/dangerous offball, and (probably most importantly) he spends more time than Kyrie in threatening positions behind the arc, which effectively makes the floor bigger. Kemba's lower 3PT% mostly comes from taking a lot more of them, and having those attempts be tougher. As we've seen with Harden and Lillard, it's absolutely worth making that tradeoff, because it makes the defense's life so much harder in terms of the ground it has to cover.
I predict that we'll see both Brown and Tatum play better with Kemba, even if just because they'll find themselves in advantageous positions more often (in addition to age-related improvement).
Now, on the defensive end, they are going to give a lot of those points back, and miss Horford (and Baynes) a ton. We all know how good he is there, and how bad Kanter is. Teams with dangerous shooters off the dribble are going to make a mockery of the Celtics. Against teams without that, like Philly, the Celtics might be surprisingly competitive on the defensive end--Kanter was less of a liability against OKC, for obvious reasons.
TLDR of my view: "replacing shots" isn't the way to look at things. Kemba can replace a ton of shots just on his own, and he's done that before to make a bad teams efficient offensively. The Celtics have a lot of plausible paths to a good offense, but way fewer to a good defense (barring a trade at the 5).