It depends on the amount of the signal and quieting slope of the receiver/tuner. See the picture below.
View attachment 1232548
Note how a small increase in RF signal can improve the quieting of the signal a fair amount.
At the lower RF level the signal to noise ratio (S/N) is only about 42 dB. This would be a somewhat noisy signal. At the higher RF level the signal to noise ratio is 50 dB and that would likely be very close to noise free.
Of course this is dependent on the quieting slope of the particular tuner/receiver being used.
This is the calculation for cross polarization signal loss, for any angle of cross polarization.
Loss (in DB) = 10*(log base 10)(cos(theta))^2
I am not sure how one comes up with 3 dB loss. That was not the correct answer when I took the test for my first class radiotelephone operator’s license.
In theory the loss is infinite, in the real world it can approach and exceed 20 dB.
That said, if one looks at the FCC license data for FM broadcast stations, one will find that a lot of them broadcast the same amount of power in the horizontal and vertical planes. In fact circular polarization is not that uncommon.
It is not uncommon for the FM signal to change polarization on its trip from the transmitter to the receive antenna, due to reflections and so on, so it is not a bad idea if one is having trouble receiving a stations of interest to change the orientation of the antenna.
Below is a generic picture that shows what happens to the radiation pattern of a dipole antenna when a conductor (the shield of the feed line coaxial cable for example) is placed in the RF near field.
Note the distortion of the pattern and the rather deep nulls (directions in which the antenna's reception is reduced).
Because of the variables involved in each installation it is somewhat difficult to predict the exact RF pattern.
View attachment 1232666
This is an example of what can happen if one just drops the coaxial cable next to and parallel with the element the dipole antenna. Doing this may also change the tuned frequency of the dipole antenna. This will happen even if one has taken steps to reduce the unwanted common mode current on the outside of the shield of the antenna.
BTW, 3 ferrite common mode RF chokes are very likely to be better that the coiled coaxial cable choke across the full FM broadcast band. Just connect it to your RF VNA (vector network analyzer) to see the difference.
There are now over 500 hits on this thread and to those that are following this thread, one can't just say that all of the signals that those following this thread may have will be powerful enough, that is have enough RF field strength that paying attention to some of the small details will not help.
Again, this is just basic text book RF 101 stuff...The stuff that can make or break the reception of weaker stations. There is nothing wrong with just kludging together an antenna, but knowing about and paying attention to the small details can remove a lot of uncertainty from the antenna build and in doing so, possibly, if not likely, improve the performance of the antenna.
And to musichal, just build your copper pipe dipole antenna, use 3 ferrite RF common mode chokes at the antenna feed point and run the coaxial cable at right angles from the dipole for as far as you can, at least a half wave length if you can. This will make your antenna perform about the best that it can.
If one does this they can be fairly confident that their antenna is working about as well as it can for the amount of RF field strength (radio station signal strength) that is available to the antenna.
.