Vertex count vs. Dynamic batching

I'm working on a game with procedurally generated voxel enemies. I made a test mesh to see what the performance was like. I quickly noticed that i was getting 1 draw call/mesh even with a fairly simple mesh.

I read about Dynamic batching rules here and here. Based on this info I figured the problem must be because of either the shader I'm using and/or the Vertex count. I cut my generated mesh from 440 vertices to under 300 and TADDAA, 1 draw call/mesh.

The problem is I have no idea why. I'm using as simple a shader as I could find in Unity's Shader Tutorials. And if I generate another kind of mesh with 900 vertices, I can get that to be batched without problems.

So mesh A with 900 vertices and Material M does get batched but in the same scene Mesh B with 440 vertices and Material M doesn't get batched.

I made a test app to try this out.

If I use "option A", to create a 900 vert, 100 triangle mesh with my simple shader, and generate 10 of these objects, I'm seeing 1 draw call. (why does Unity say 100 triangles when it should have 300 ?.. anyways...)

If I use "option B" to create 10 of my 440 vert Meshes, i get 10 draw calls (same material)

Code (CSharp):

1.     public static Mesh _Mesh;
2.     public static Material _Mat;
3.  
4.     void Awake()
5.     {
6.         if (_Mat == null) _Mat = Resources.Load<Material>("TestMat");
7.  
8. //        option A
9. //        if (_Mesh == null) _Mesh = createMesh();
10.  
11. //        option B
12.         if (_Mesh == null) _Mesh = Resources.Load<Mesh>("test1_Mesh");
13.  
14.         MeshRenderer mr = gameObject.AddComponent<MeshRenderer>();
15.         mr.receiveShadows = false;
16.         mr.castShadows = false;
17.         MeshFilter mf = gameObject.AddComponent<MeshFilter> ();
18.  
19.         mf.sharedMesh = _Mesh;
20.         mr.sharedMaterial = _Mat;
21.     }
22.  
23.     Mesh createMesh()
24.     {
25.         int vertCount = 300 * 3;
26.  
27.         Vector3[] verts = new Vector3[vertCount];
28.         for (int i = 0; i < vertCount; i+=3)
29.         {
30.             verts [i] = new Vector3 (0, 10, i);
31.             verts [i+1] = new Vector3 (10, 0, i);
32.             verts [i+2] = new Vector3 (-10, 0, i);
33.         }
34.         int[] tris = new int[vertCount / 3];
35.         for (int i = 0; i < tris.Length; i++)
36.         {
37.             tris[i] = i;
38.         }
39.  
40.         Mesh m = new Mesh();
41.         m.Clear ();
42.         m.vertices = verts;
43.         m.triangles = tris;
44.         m.RecalculateNormals ();
45.         m.Optimize ();
46.         return m;
47.     }

What gives ? Is there more to the batching than the docs tell ?
My test case should rule out any other variables since nothing is scaled, there's no lighting or shadows etc.
Am I just having a brain fart...

Shader I'm using in both cases.

Code (CSharp):

1. Shader "VertexLitTest" {
2.     Properties {
3.         _Color ("Main Color", Color) = (1,1,1,0.5)
4.     }
5.     SubShader {
6.         Pass {
7.             Material {
8.                 Diffuse [_Color]
9.             }
10.             Lighting Off
11.         }
12.     }
13. }
14.