What speaks against doing this?:
public struct T { private float[] Elements { get; set; } public T(params float[] elements) { Elements = elem