Can I wind my 9:1 unun with two wires instead of three?

All the 9:1 unun tutorials out there are using three wires to wind, but I was wondering if it is possible to wind it using two wires similar to how this 49:1 is built (https://elginradio.files.wordpress.com/2017/10/49-1-unun.jpg). Of course, I'd modify the ratio to be 1:3 instead of 1:7.

Does anyone see any issues with this?

For illustration, here's a 9:1 unun construction using 3 wires from M0UKD:

Since the A and B windings are connected together in series, there's no compelling reason you couldn't replace them with one longer wire that has the same number of turns.

For an ideal transformer only the number of turns would matter, and how those turns are arranged would be irrelevant. In practice how the turns are wound affects things like the parasitic capacitance: this is why you can "tune" an inductor in a filter a little by pinching or spreading the turns. Whether this change makes the device better or worse for your application will depend on details like the frequency, core dimensions, wire gauge, and so on.

I've read that crossing over the middle of the core as in your question is supposed to reduce the capacitance between the input and the output. My guess is it makes negligible difference in practice. I like that it makes the wires come out on opposite sides of the box and it looks neat. Easiest way to find out is to build it and measure empirically.

Using triple windings allows closer wire spacing between the transformer input and output (primary and secondary) windings which can improve the magnetic coupling coefficient. This can slightly improve power transfer efficiency, as more power goes into the output/antenna windings instead of heating the toroid. This possibility of core heating can make a difference at higher QRO power levels, as any heating of the ferrite material may affect permeability and further reduce flux density. But this may or may not be measurable at QRP power levels.

There are other geometries of overlapped 2 wire windings used in some transformers (triangular overlaps of Litz wire on ferrite bar antennas, etc.), but on a toroid they probably end up looking similar to the canonical 3 wire winding.