So my thought was to modify the headers 4 into 1 and back into 4. But only do that on one side.
Boy . . . . . I'd have to pass on that idea.

That is if this were any kind of performance related intention. and where the headers 'were' suppose to provide any function toward that end.

One problem you'll probably suffer is that the O2 sensors are intended to be close enough to the exhaust exit of the cylinder to be heated, too far down line (or even possibly just "sticking out in the wind"

) and they're running too cold and won't necessarily provide an accurate reading value; and even if so, then they might not be getting hot enough to cook-off the by-products of combustion that will likely tend to adhere to them, this also causing inaccurate values and a shorter useful service life span.

Is this need for an O2 sensor because your adopting an "E.F.I." system? Or just for the purpose of utilizing a "lean-rich"/"air-fuel ratio" instrument?

Particularly if the latter, all I can say is: It ain't worth it! Better to just "read the plugs".

And then, as previously mentioned, the exhaust must be "sealed-up", no leaks; but even then, with the side pipes presenting a rather short distance length from the proposed O2 sensor and the open atmosphere, if with a sum of camshaft overlap, or any other reason causing significant reversion in the pipes, the functionality in the sensors is lost!

Scott.