So, I've been doing some number crunching, and, based on what I've come up with, I'm really not sure the whole "Stamkos doesn't mess with the Leafs' cap" argument holds water.
With Marner, Matthews, Nylander, Soshnikov, Zaitsev and Hyman on the opening day roster, $750K 1 year contracts for the 3 RFA defencemen, $1M, 1 year for a backup goalie, the Cowen buyout credit, no performance bonuses included, and a rough estimate of a $73.2M cap (current cap with 2.5% growth - which was the last estimate I saw of what it might be after the PA uses the escalator) I put the Leafs at ~$7.8M in cap before they get into LTIR space.
Now, why is the pre-LTIR number important? Because it can't be used to cover performance bonuses. Unused LTIR space disappears at the end of the season - and, if you use any of it, you're considered to have hit the cap ceiling - and performance bonuses are awarded after. Depending on what Schedule B bonuses are in Matthews' contract, the Leafs will have somewhere between $3.6M and $5.6M in available bonuses on the roster next season - and, if they go into LTIR space, every penny of those bonuses will be carried over into the 17/18. Now, obviously, it's unlikely every one of those bonuses are reached, but, I'd say it's not unreasonable to estimate that ~$2M of them will be.
So, with that, in 17/18, the Leafs would be looking at a total of roughly $4.5M in dead space, while still having a number of contracts that will require space for bonuses. That puts them at ~$46.75M committed (or ~$26.5M in space) to 9 forwards, 2 defenceman, and no goalies - still with $2.75M+ in potential bonuses (depending on whether or not other prospects make the roster), Zaitsev and Hyman needing new contracts (as well as the 3 RFA defencemen), and uncertain growth in the cap. That's a pretty tight squeeze, as far as I'm concerned.