I'm not too strong on this for several reasons. Even shotguns can miss with shot. LRMs and artillery are often used indirectly, which is easy to miss with by not even hitting the same "hex" they target. Streaming fire weapons, like the HAG, also can easily miss.
There is a reason the ability to miss is put in to the system. We're not talking about infantry running in to a flamethrower, but using indirect and shot weapons to hit large targets with who are rather far away.
Yes it would speed things up to not have to track hitting, but one could say that about any weapon system, especially with the "advanced" systems of the 31st Century. So, I'm going to say no, due to it losing out on part of the granularity that Battletech provides as opposed to how abstract 40K is. Even if the Cluster table was set up to include a fair amount of "zero" in the cluster table, I still think it would be a bad thing to introduce.
But still, even with my idea, it doesn't hit by a frag all the times, and actually most of them are turned to zero as you have pointed out. For example, a +9(assumed that a regular gunner, walks, shoots against the enemy that moves 5 hexes, that is in the medium range bracket) results -6 in the cluster roll, makes 7 or less(which is about 58%) results no hits at all, and only scores the minimum fragments on 8+. A +9 to hit is hit by around 27.7%, and the penalized cluster roll will represent the difficulty to hit well, I think. Even with this the chance to actually damaging the target is not so weirdly different with a to-hit roll, so I think that it still represent the rate of miss enough.
You cannot hit with a full salvo with the lucky roll with this, and it can be a problem. But is there anyone that seriously expects all the salvos are actually hits? SRM and LRM are shoot in groups rather than one by one for a reason - to make the target harder to evade the attack. SRM and LRM, and LB-X are not the streak SRM. Streak SRM requires the mark and all the missiles are hits. But SRM, LRM, and LB-X are designed to be spread and pray to hit at least some of them. That's why I think that it will better represent the cluster weapons of real Battletech worlds. Roll for the hit then also roll for the independent cluster roll is more suited for the Streak SRM's trait, and it does not represent the actual clusters correctly.
It doesn't makes the game like 40k. Actually, attack process in 40k have much more steps, and they use 1d6 per a roll. But Battletech is different - 2d6 is the basis, and you are required to roll once per a case. The idea does not advocate of 1d6s either(and I hate 1d6 to determine something and I prefer 2d6 or more).
Technically the number of missile hits (as is) is based on a 4d6 roll - 2d6 to hit, then 2d6 cluster roll... But as it's split in two, any attempt to combine them into one roll will end up skewing the result. As I wanted to skew the results 3d6 wasn't a problem for me, but I don't really know which alternative is "worst".
No I didn't mean like that. That's because the expected result of 2d6 and 3d6 are different.
2d6 may results 11 possible results, which are;
2 - 1/36
3 - 2/36
4 - 3/36
5 - 4/36
6 - 5/36
7 - 6/36
8 - 5/36
9 - 4/36
10 - 3/36
11 - 2/36
12 - 1/36
And 3d6 may results...
3 - 1/216
4 - 3/216
5 - 6/216
6 - 10/216
7 - 15/216
8 - 21/216
9 - 25/216
10 - 27/216
11 - 27/216
12 - 25/216
13 - 21/216
14 - 15/216
15 - 10/216
16 - 6/216
17 - 3/216
18 - 1/216
It means, bring 3d6 instead of 2d6, with the chart that is expected to be use with 2d6, will cause much problem than just a flat number of add or subtract. In my idea, still a total modifier of +3 makes the roll just same as the normal to-hit modifier, although it rarely happens. Perhaps it would be possible if the target is immobile and get a -4 to the modifier?
That's why a 4d6 and two separated 2d6s are totally different, if you consider the chance to get a result.