"Anyway, the code does not quantify the distance with rs232 may be because the authors didn't know of cable qualities in the future."

Nope.

The spec doesn't specify cable length because it is not the cable length per se that is the limiting factor.
The actual limiting factor is the load presented to the driver by the cable - principally the capacitance. It is irrelevant whether you have a short length of high-capacitance cable, or a long length of low-capacitance cable - it's the total capacitive load that makes the difference.
That's why the spec just states a load requirement:

"The effective shunt capacitance (C_L) of the receiver side of an interchange circuit, including the capacitance of cable, measured at the interface point, shall not exceed 2500pF."

Annex A provides guidance on the characteristics of the interconnecting cable, including an example with a receiver having 100pF input capacitance, and a 30pF/foot cable:

"Dividing 2400pF by 30pF/ft reveals that the cable can only be 80 feet long. If the user requirements call for an interconnecting cable that is 200 feet long, then the user must select a cable that has only 8pF of mutual capacitance (C_m)"

So the spec itself actually mentions 200ft - which is almost 61m!

The length limit is imposed by the cable; not by the spec!

" I think beyond this distance any improvement in the cable quality would hardly add a centimeter."

Yes, of course there is a practical limit to how far you can reduce a cable's capacitance and as a result of this there will be a practical upper limit on cable length.
But it is incorrect to say that RS232 itself has any particular length limit; if you do that, people will take it to mean that RS232 can do 50m (or whatever) over any cable...!