Header diameter on a small 4 stroke single

I may have posted about this before, but....
Things is, I have this long term Suzuki 125 project, originally meant for my youngest son. A Suzuki GN125 converted to a mini street tracker. 3.25x18 tyres at both ends, and a 70 mm shortened fork took care of levelling it. The engine got a 61 mm 11:1 Wiseco piston, taking it to 141 cc. And 2 mm oversize valves, and a 3 angle valve job. Plus a 25 mm DellOrto with a velocity stack and a K&N.
Unfortunately, I made the exhaust pipe way too big, 38 mm OD, 35.5 mm ID. So I did some research around aftermarket pipes for the DR-Z 125, which has pretty much the same engine. And they all have a stepped header, starting with 1 1/8" OD, then stepping up to 1 1/4" about half way.
When I just checked the port diameter, it is in fact 29 mm, as cast originally. So it would be a smooth transition if starting directly with a 1 1/4" OD pipe.
So what to do? A stepped 1 1/8" - 1 1/4" pipe like the DR-Z aftermarket pipes, a straight 1 1/4" pipe, or a stepped 1 1/4"- 1 3/8" pipe?

Here is my take on it .
All of those option will work, one may work better for your intended use than another. Without comparison testing or simulation software it will be difficult to know which is best for your purpose.
In this position I would choose the easiest option.

I like the stepped option best,.next single diameter 1.125 and last 1.25. I have no science to back that up but I like to err on the side of caution with intake and exhaust diameter.

A 29 mm insert in the head pipe would seem to be the easiest solution. They are sold as "anti reversion" inserts. They cost $40 or $50 for a pair for XS650s but if you made the header already, try making an insert. Or am I missing something here?

Thinking out loud and perhaps wrong
Diameter ( Area changes) causes resistance
And affects velocity and pressure

port diameter, it is in fact 29 mm
The pipe you have manufactured is ID 35.5 mm ID.
Making it 35.5 minus 29 6.5 > 3.25 radially.
Although not perfect I am thinking wort testing
I depends on overall velocity . And how close can one get it there has to be tolerance for eccentricity.
No mention of muffler
If it is not working I would try to make a cone and put inside and test that.