In `my_func(a)`, **Recurrence Relation** will be 

$T(n) =
  \begin{cases}
    4T\bigg(\frac{n}{2}\bigg)+{n} & \quad \text{if } n \geq 4\\   
    1  & \quad \text{if } n <4
  \end{cases}
$


In `new_func(a)`, **Recurrence Relation** will be 

$T(n) =
  \begin{cases}
    3T\bigg(\frac{n}{2}\bigg)+{n} & \quad \text{if } n \geq 4\\   
    1  & \quad \text{if } n <4
  \end{cases}
$  

You can solve Both of these **Recurrence Relations** using [**Master Theorem**](https://ediscretestructures.blogspot.com/2022/03/master-theorem-for-recurrence-relation.html) as explained in link.  

The **Time Complexity** of `my_func(a)` will be $\theta(n^2)$ since $\log_24 = 2$  

The **Time Complexity** of `new_func(a)` will be $\theta(n^{1.5849})$ since $\log_23 = 1.5849$  


You can solve both of these questions by **Substitution Method**, which is Time Consuming. One of the [Example](https://cs.stackexchange.com/a/143992/143377) using this method is attached.


The `new_func(a)` will be slower than `Merge Sort`, and faster than `my_func(a)`.