It's an interesting idea, but possibly creates a balance issue, either your teams take full advantage of the researchers skill and allow you to absolutely speed through 1 or 2 research projects in lightning speed, or you could make additional researchers added to a team only add a minor bonus, in which case there's an issue where you'll have researcher shortages.
I'm of the opinion that if this was added additional researchers should only add 10% of their bonus to the project, and total number of labs should be limited by what the leader can handle.
It does add extra micro management though, for limited advantage, but you'll get greater training if researchers.
There's lots of ways to go with this, but the way I had in mind would have no balance impact at all. To illustrate the point, here's a typical situation from one of my games:
Alice Jones, 35% DS, with 20 labs working on Armor
Bob Smith, 20% MK, with 20 labs working on Warhead
Carol Davis, 20% DS, with 0 labs working on Genetic Mod +20C
Dave Smith, 15% MK, with 0 labs working on some kind of railgun project I don't care about
Ed Jones, 40% CP, with 0 labs working on some kind of particle beams project I don't care about
[20 other researchers, of varying skills, all with 0 labs, working on projects that I don't care about]
- If Alice dies, I cancel Carol's project, then restart the Armor project with Carol running it, since she's my next-best DS researcher.
- If a new researcher shows up, I take a lab away from Bob, look around for some project that doesn't interest me, assign it to New Researcher with 1 lab, then take that lab away and give it back to Bob.
- From time to time I look over the researchers to see if anyone's overtaken my current leaders as the best available.
Here's how I imagine it might go instead.
The DS Team [Alice 35% DS, Carol 20% DS, and two other DS researchers] working on armor with 20 labs. Team effectiveness is 35%, the bonus for the best researcher on the team.
The MK Team [Bob 20% MK, Dave 15% MK, and one other MK researcher] working on warhead with 20 labs. Team effectiveness is 20%.
[several other teams, none doing anything interesting.]
- If Alice dies, nothing happens. There's a message, but no interrupt. I do nothing. The project continues with Carol's bonus of 20%.
- If a new researcher shows up, I just join them to one of the existing teams.
- I never review the projects to make sure that the best researchers are on the job. The nature of the teams is such that if one of the researchers overtakes the previous best, they'll naturally slide into the leadership role.
The net result of all this is the same bonuses applied to the same labs on the same projects. It just minimizes the busywork involved with keeping all your researchers busy so that they skill up properly.
This cat might not even need skinning, if idle researchers skill up the same way that working researchers do.