Clarify the Matrix3x2 docs to indicate that it is supposed to be used for 2d operations.

The current docs for Matrix3x2 have little to no "leading" info or remarks about the type. I would love to see here that the Matrix3x2 is meant to be used for 2d operations, and that this matrix is actually a 3x3 matrix of which the last column is always [0 0 1].

In [the remarks](https://github.com/dotnet/runtime/blob/f8218f9c4aaf4d6725524dc5d9e14c985e0f1b51/src/libraries/System.Private.CoreLib/src/System/Numerics/Matrix3x2.cs#L327) of the `GetDeterminant` method, it is stated that the method treats the matrix as a 3x3, but the way the remark is made, it seems that this only applies to the `GetDeterminant` method.

When working in 2d space, you can chain multiple matrix operations together into a single matrix, by multiplying the matrixes. Multiplying matrixes is therefore a very important operation, so when I took a look at the source of the multiply method, it showed me that this code also treats the matrix as a 3x3 matrix with the last column as [0 0 1].

If the Matrix3x2 was truly a 3x2 matrix, the [operator *](https://github.com/dotnet/runtime/blob/f8218f9c4aaf4d6725524dc5d9e14c985e0f1b51/src/libraries/System.Private.CoreLib/src/System/Numerics/Matrix3x2.Impl.cs#L145) should (NO NO NO, I AM NOT SUGGESTING WE DO THIS XD) instead be:

```diff
[MethodImpl(MethodImplOptions.AggressiveInlining)]
public static Impl operator *(in Impl left, in Impl right)
{
Impl result;

result.X = new Vector2(
left.X.X * right.X.X + left.X.Y * right.Y.X,
left.X.X * right.X.Y + left.X.Y * right.Y.Y
);
result.Y = new Vector2(
left.Y.X * right.X.X + left.Y.Y * right.Y.X,
left.Y.X * right.X.Y + left.Y.Y * right.Y.Y
);
result.Z = new Vector2(
- left.Z.X * right.X.X + left.Z.Y * right.Y.X + right.Z.X,
- left.Z.X * right.X.Y + left.Z.Y * right.Y.Y + right.Z.Y
+ left.Z.X * right.X.X + left.Z.Y * right.Y.X,
+ left.Z.X * right.X.Y + left.Z.Y * right.Y.Y
);

return result;
}
```

But this would invalidate all the use cases for 2d space, since the calculations for multiplying matrixes for 2d use are then incorrect.

So that is why I would love to see some documentation indicating that this is what the class should be used for, and a general remark that the Matrix works as a 3x3 matrix with a default last column of [0 0 1].

I would like to hear if I understood the code & it's use correctly, or what the thinking here is. If need be, I can create a PR addressing as such in the near future.