Ideally through the draft. If through trades for our best pieces, let's get players who are young and have longer futures with us. Larsson is 26 already and despite the praise you're giving him, isn't all that special a player imo. I'm not sure how much offense he has in him. He hasn't broken 20 points in almost 5 years, that's not a high ceiling for a d-man with offense. Puljujärvi is a totally different thing than Larsson, his age is right at least. But he's a risk and definitely not one worth moving a Kreider at his absolute peak for.
I am going with headlines since this post covers two topics and will be long as usual...

The blue-line should be built before a group of forwards is assembled
I can't get that approch to add up. We draft a nuke forward this summer. 2019. 2020 and 2021 we take nuke Ds. Ds aren't carrying their teams by the time they are 22 y/o. It takes more than 4 years. 5-6 normally, if not a year or two longer before a young D is really ready to be a goto No 1 guy. Then they have longer careers. If we count on the kid we drafting getting it done super fast -- in 5 years -- the first guy is ready in 2025 and the second in 2026.
In almost all cases, if we look at successful rebuilds, its younger forwards back-up'ed by a more experienced blueline. Right? I think there is a reason for that. The other way around can be very problematic. A blueline is so important for a team, it must be steady. And you can only get there by going through going pains. You don't want those young forwards first to go through a couple of tough years when they break into the league and then -- in addition -- go through 4-5 more tough years when a blueline is built under them. The result just isn't going to be good.
Duncan Keith and Brent Seabrook was 24-25 when Chicago drafted Pat Kane. That is perfect age from my POV.
D age is forward years minus 3, at the very least
Its also important to remember that Ds won't start to break down when they are 28-29 like forwards. All over the league we have Ds playing into their mid-30s. I would at least say that D years should be counted as forward years minus 3.
Ie a defensemen age 26 is 100% comparable to a forward aged 23, in general of course. If a forward starts to trend downwards around 29 these days on average maybe, the same age is 32 for a D.
If we look at the new generation of Ds that have entered the league after 05', there is no way these guys will be out of the NHL at the age of 31-32.
Burns is leading Ds in scoring and he is 34. Giordano is 2nd in D scoring and he is 36. Letang 32. Yandle 33. 6th and 8th in D scoring. Suter us 34. Buff is 34. There are sooo many good Ds coming up behind that generation, Carlson, Karlsson, Doughty and co. They are 29-30 now, no way they won't be around playing top minutes for another 4-5 years at the very least. Many of them past that too probably. Also depends on how much hockey you play, those long PO runs wears on players of course.
So I don't for a second believe that an Adam Larsson wouldn't be able to play a good 8 more years in this league, with a slight downward trend staring in year 7-8...
I get where you are coming from, I would be totally against getting a Vlad Tarasenko for the same reason, but I wouldn't apply it to a D that turned 26 two months ago. Like its not even unlikely that a top D that is 26 today will last longer than like a forward like Brett Howden that is 21. Many forwards tackle off in their late 20s.