At 5-on-5, Winnipeg uses him as an offensive #4 defenseman, giving him #4 minutes, #5 level competition, and starting him more in the offensive zone than the team does without him.
Over the past two years, his goals for percentage is 5.09% less than the Jets' goals for percentage without him, and his expected goals for percentage is 2.41% lower than the Jets without him. In both cases, they are below 50% with him on the ice, and above 50% without him on the ice. The same is true for every single metric - corsi for, fenwick for, high danger chances, etc. Jets win the battle with Myers off the ice and they lose the battle with him off the ice.
It's worth noting that Dmitry Kulikov, by far his most common defense partner over that time frame, is bad. Myers does slightly better without Kulikov, but Kulikov also does slightly better without Myers, and the difference in all cases is not huge. Even when excluding Kulikov in both cases - comparing Myers without Kulikov to his team without Kulikov - the team still does better without him on the ice. His most common forward linemates - Bryan Little and Patrik Laine - are also not good. He does much better away from Bryan Little in particular, but Little also does better without him.
His 5-on-5 production is also nothing special, both in terms of scoring rates and aggregate production. It's solid - good enough for a top-pairing guy if he were good at everything else. But he's not, so it doesn't make up for it.
I'm sure that he would do better with a superior defense partner, but every defenseman in the league would do better with a better defense partner. Myers' 5-on-5 performance, even after accounting for the deficiencies of his defense partner, is just not good. I don't think you'll ever get truly good results out of him unless you use him in a #5 or #6 role.
On the power play, he is pretty good. His power play scoring rates over the past few years are elite and his shot impacts on the power play are solid. The Jets do score goals and generate expected goals more frequently than they do without him, but that is because of who he plays with. He plays almost all of his PP minutes without Schiefele and Wheeler and they both do much better in the minuscule sample size that they do play with him. A new team that acquires him could probably stand to use him on the first PP unit, but I don't think he could be a top-unit PPQB. More of a complementary player in that role. However, I haven't seen quite enough of him on the PP to make that distinction.
Both of the wins above replacement models I've looked at rank him as a below replacement level player.
He's a #5 defenseman. He generates enough offense to drive a third pairing, but he is poor defensively and his defensive shortcomings will out-weigh his offensive strengths if he is in the top-4.