6 knights vs 2 queens

I'd expect it to be similar to 3 knights v 1 queen.
In that ending 60+% of positions with the knights to play should be won by the knights and around 25% of positions with the queen to play should be won be the queen. Just under half of all positions should be drawn.
In KNNNNNNKQQ there are good chances for the two queens to swap themselves for four of the knights leaving a drawn KNNK ending.
The ending is quite unlikely to occur in practice.
5 queens vs 10 knights can be a draw. I tested this once in stockfish.
That ending would probably be even less likely to occur in practice, but I don't think you could get a definitive answer from Stockfish.

4 queens vs 10 knights is a draw.
I have played this many times on both sides...oh wait...no I haven't and neither has anyone who has ever played a game of chess (nothing personal ES, the same applies to the OP).

Starting with a reasonably balanced position (eval +1.0 favoring white) and letting Stockfish8 play both sides while analyzing about 12 moves deep (23 ply) gives the following game. This is not definitive, of course, but interesting, shows the dominance of the knights if they protect each other and concludes in one of the most aesthetic mating patterns you will ever see.

4 queens vs 10 knights is a draw.
Prove it.
I tested it on my program.
Starting with a reasonably balanced position (eval +1.0 favoring white) and letting Stockfish8 play both sides while analyzing about 12 moves deep (23 ply) gives the following game. This is not definitive, of course, but interesting, shows the dominance of the knights if they protect each other and concludes in one of the most aesthetic mating patterns you will ever see.
As you say not conclusive, but consistent with what I said in #3.
4 queens vs 10 knights is a draw.
Prove it.
I tested it on my program.
Try your program on this position. I bet it can't draw with the knights.
4 queens vs 10 knights is a draw.
Prove it.
I tested it on my program.
Try your program on this position. I bet it can't draw with the knights.
Do you mean the Black king is on e1 and the White king is on e8? The Black king already seems to be already checkmated in the diagram.
Exactly. I've noticed it's usually very difficult for a player in a checkmate position to force a draw.

Getting a little far out with 4Q v 10N but Stockfish demonstrates knight dominance, again starting from a reasonable position with no easy forks for white. There is also a remarkable mate involving three consecutive sacrifices by white with black moving ...KxN.

The five queens defeat the 10 N but it takes nearly 40 moves to mate. Very glad I don't have to play this position myself. My brain would need weeks to recover.

Multiple queens is more redundant than multiple knights because major pieces gets devalued in the presence of lower valued minor pieces.
Larry Kaufman called it «Principle of the redundancy of major pieces»
"During the analysis, I discovered something very remarkable: the board is simply too small for two Queens of the same color. They only get in each other's way. I realize that this might sound stupid, but I fully mean it. The advantage is much less than one would expect by counting material." - Viktor Korchnoi
Stockfish does not have penalty for multiple queens. Komodo does, thus i recommend to use the latter for such positions.
Analysis by Komodo 11.2.2 64-bit:
1.Nfe4 Qf8 2.Ncd4 Qca6 3.Ncb5 Qad7 4.Nfg5 Qhh8
+/- (1.37 ++) Depth: 24 00:00:35 663MN, tb=73706
(08.11.2017)