Is this taking into account the time needed to slow down?
it isn't realistic assumption. Until you're talking pure solar, the amount of acceleration is limited by the reaction mass available. Actually to get there in 10 years with the Isp 3500 3 stages are necessary, or better the Isp should be increased 2x-4x - still seems doable - to get with like 2 stages with realistic [today] parameters of the reactors/etc.
Compare to nuclear powered ion thruster. Say we get reactor plus generation at 5KW/kg total - takes some engineering, yet nothing unrealistic for current tech (even 10KW/kg seems pretty reachable). Reactor is on a long pole with only small protection wall directly between reactor and payload. Say 5 ton reactor, 25MW. 100 ton whole rocket, 80 ton of it reaction mass. At current NASA 40km/s ion trusters we get delta-v 80km/s in 60 days. If we get thrusters with 80km/s - wikipedia mentions that current ones reach 50km/s, so don't see why we can't increase voltage and thus ejection speed further - then it would take 240 days to reach delta-v 160km/s (i.e. current multi-year missions to Jupiter/etc. would get in well under a year, and it will be with like 10 ton payloads). Don't see solar sails coming close to that - https://en.wikipedia.org/wiki/Solar_sail#Inner_planets.
And as i mentioned earlier - let say we got thrusters with 400km/s. The same rocket will get to 800km/s - 1500 years to the nearest star - in 20 years. 3 stages - 500 years to the nearest star. 1 ton final payload if starting with 1000 ton rocket like the one described above.
Gathering reaction mass ram style - it needs big apparatus and needs to be efficient. Doesn't seem realistic with current tech, yet i'm sure will be on the table once the tech matures.
2. 1 ton is starting from 1000 ton rocket. The Saturn V and Startship are on the scale of 5000 ton and assembled on Earth. That interstellar rocket will be assembled in space anyway, so not being subject to any meaningful gravitational nor accelerational stresses, we can easily build a 100000 or even 500000 ton rocket - basically just the reactors and tanks of acceleration mass - and thus get 100-500 ton payload. If we get any [semi]hybernation going (may be combined with 3d printing or CRISPR-like repair of organs, whatever we get in 20-30 years) or more probably some brain [partial] upload integrated with AI into some capable cyborg, may be even some people or those merged human/AI cyborgs would be able to go.
And by collecting some additional reaction mass ram-style over that distance and time (as long as we have enough reactor power to use part of the collected mass to avoid slow down resulting from the collection) we'd probably be able to slow down some small probes to land and orbit various objects in the target star system.