Analytics only go back to 2007-2008, but Holmstrom was a fine producer with terrific analytics at 35, a great producer with good analytics at 36 and maintained those analytics but with massively decreasing production after 36. If you think Hornqvist is going to age the exact same as Holmstrom, Hornqvist has like 3 more years of being an effective player before his production falls off a cliff at ES.
But I don't think Hornqvist at 34 is the same at Holmstrom at 34, though. Hornqvist in the last 2 years from age 32-33 is more like Holmstrom from age 34-35. Hornqvist in the last 2 years has a 1.89 5v5 points/60 with a +2.63% xGF% relTM. Holmstrom from 2007-2009 was at 1.95 5v5 points/60 with a +6.96% xGF% relTM. Holmstrom fell off a cliff with his production after 2009, where he went from 1.95 in 07-09 to 1.51 in 09-10 to 1.21 in 10-11 to 0.75 in 11-12. His analytics stayed good (outside of 11-12, where he was totally done as a player), but his production completely slipped.
The point I'm making here is trade Hornqvist before you can't trade him anymore. Kessel last season was moved for the same reason, he's a ticking time bomb for when he falls off a cliff and you don't want to be holding it when it goes off. Hornqvist has at most 1-3 years left of being a solid contributor, and considering he doesn't fit in the top-6 and Sullivan isn't married to him being on the top powerplay unit, I genuinely see no reason to keep him.
Edit: to add, Hornqvist went the same route that Kunitz did. Kunitz's production started slipping in 2014-2015, when he was 35. He stayed an effective top-6 player (albeit overplayed) for 2 years, but then fell into a bottom-6 player. If you want to apply that standard to Hornqvist, it's the same 1-3 years.