I think part of the reason that the leafs are having to play their young defenders such limited usage rather than keeping them in the AHL is because they need their ELC as the cap is so tight now.
I don't think this was true. The Leafs' picked up several league minimum D during the summer (ie Harpur, Schmaltz etc) specifically so that they could keep Liljegren and Sandin in the minors. That is what Dubas said at that the time. Those are good candidates for sheltered, low-minute D. Sandin and Liljegren were not.
And 900K ELC Ds on the third pairing are not good deal. 900K ELC Ds in the top-4 are a great deal.
Also don't think using a top 5 pick like Rielly is the best example as top 5 picks usually hit the NHL at an earlier age and play more time. Same goes for guys like Hughes, Makar, Dahlin etc. Just a different tier of defenders to Sandin and Lily.
Age isn't really the most important factor here. The reason I used Gardiner, Rielly and Dermott is because those really the only D under the age of 22 that the team has brought into the NHL in decade. The case would have been better for Rielly getting low minutes in the NHL at 19 because the Leafs couldn't send him to the AHL. Both Lilly and Sandin could have spent the entire year in the AHL. It was not like either were simply too good to play in the AHL. I watched them both play a number of games in the AHL this season, and sometimes they looked really good, but on the other hand as I wrote at the time for one two game stretch I attended in either late November or early December, the two were possibly the worst two D on the ice those two games - for either team.
So they could have kept them out of the NHL until they were ready to fill a major role and when the team was ready to do it right. Duncan Keith was 22 (with 154 regular season AHL games) before he played his first game in the NHL - and was Chicago' #1 instantly. Sandin has 65 regular season AHL games and Liljegren 127.
On the other side of things McAvoy was 14OA. Started playing in the NHL - in the playoffs - at a younger age than either Rielly or Sandin - and averaged more than 26 minutes a night during those playoffs. Jokiharju was taken 29th OA - started playing in the NHL at a younger age than either Rielly or Sandin - averaged more than 20 minutes a game over his first 11 games.
But it is also not just about ice-time - it is also about who they are playing with - as you also mention.
The leafs havent ever had a proper partner for Rielly (or even Gardiner for that matter) let alone for any of the young guys coming through. This is really all due to the cap constraints and it will be even more difficult to bring in proper partner for Lily and Sandin. I honestly think they will be playing together on the bottom pairing next year.
Hainsey was a good partner for Rielly - yes, towards the end he couldn't handle the minutes, but he still was a very good teaching and calming presence - and he would have been a very good partner for either Liljegren or Sandin on the third pairing this season. There were other possibilities that the Leafs' could have brought in during the season (Dillon, Del Zotto, etc) - cheap, experienced, physical and defensively sound D, but they opted not too.
Both the D I mentioned were left-handed, but with Rielly and Muzzin under contract for years to come the Leafs need to either shift Sandin (or Rielly, or Muzzin) to the right side or trade Sandin (or Rielly or Muzzin) within the next year to 18 months to a team that will not block his movement into the top-4.
Yes, the cap situation that management as well as their star players have put the team in is a major problem, but the team also seems to have this bizarre need to hold on to a couple dozen NHL capable wingers while they don't have any decent D depth.
If Lilly and Sandin play most of their minutes together on the 3rd pairing next year it will be a tragedy for their development. Not only because they shouldn't be playing together, but because over the last several years Toronto plays their 3rd pairing fewer minutes and more sheltered minutes than almost any other team. There is no benefit to drafting good D if you are not going to develop them properly. I don't think that there is an easy to work around this as I assume there is almost no circumstance in which one of their prohibitively expensive contracts is getting traded. Personally, if I ran the team I would trading Kapanen, Kerfoot, Johnsson and at least one of Dermott or Holl in the hopes of freeing up cap space and getting one of the D the team was rumoured to be after (Manson, Risto etc).