It seems like DA prefers to go the route of "get the contract off the books completely" rather than have dead cap space on the books because the latter means there's less room to put toward the current roster. A contract that counts on the cap is theoretically movable in a deal; a buyout incurs dead cap space and it's not movable under any circumstance. You look at the couple deals where we moved significant contracts in full. Philadelphia, that cost us a 1st to unload Lehtera. Buffalo, that cost us writing the check for ROR's signing bonus and maybe something else that went back [Thompson or the 2nd round pick]. Both of those were apparently more palpable than having Lehtera counting ~$1.65 million on the cap for 4 years or say Sobotka counting ~$1.05 million on the cap for 4 years.
I don't necessarily agree with that approach, I'm just saying I think that's the mindset behind it.
I vehemently disagree that the value/cost for moving Lehtera was a 1st. There is absolutely no way that the draft-time value of of Brayden Schenn (with 3 years remaining at $5.125 mil AAV) was just a single mid-late 1st round pick. At the time of that trade, Schenn was (conservatively) a 25 years old, 50+ point winger who was good for 175 hits a year and well known for being able to stay healthy. That type of player fetches a 1st as a pure rental at the deadline, not over the summer with 3 years left of team control at a reasonable cap hit. Without including Lehtera, Schenn costs a 1st and some other mid-value asset. The cost to shed Lehtera's deal was the difference between that asset and a 1st round pick. Honestly, I'd wager that Philly shopped Schenn alone for two 1sts (or a 1st and a good-but-not elite prospect) and the Blues said we'll match that price, but you need to take Lehtera's contract to make the money work.
As for Sobotka/Bergie, I really disagree that we added value for them to take those contracts. ROR was getting traded by the end of the day, that's pretty clear. There were numerous reports that Buffalo didn't want to pay that bonus and July 1 is one of the busiest days of the year for an NHL GM. The fact that ROR was traded that evening makes it clear that Buffalo's front office was allocating a good chunk of their time/resources to getting a deal done before having to pay a bonus. Paying that bonus was the cost of acquiring ROR. I don't believe for a 2nd that waiting a day and trading a 1st, 2nd and Thompson without those contracts included was ever an option. The vast majority of non-rental-trades for a $7+ mil player involve money going back the other way, as that is just the reality of a cap league. I doubt Buffalo had any offers that didn't include taking salary back and I don't think ROR's value (a clear cut 1C with 5 years of team control) was much less than a 1st, 2nd and Thompson.
As a whole, I think most fans overestimate the negative value of bad contracts that are limited in remaining term and aren't for borderline star player money. Contracts like Lucic, Seabrook, Backes, Eriksson, etc are incredibly difficult to move. But $3-4 mil deals with a couple years of remaining term for serviceable middle 6 or bottom 4 guys are pretty easy to include in a trade or shed for a low-mid value asset.