OK, well, that's a nice little antenna, but why did you buy a right-angle style? My intention, when I suggested it before, was that it stick out the side of the keg in line with the connector, not at right angles to it. This was for electrical and cosmetic reasons. The electrical reason is a bit complicated, and there are two main factors. One is that, since this is a monopole antenna, it needs a ground plane to work optimally and the ground plane must be at right angles to the whip (ie. across its bottom). Also, as we are working at a very high frequency, the distance from the base of the black part to the ground plane should be minimized. These things affect both the impedance matching (discussed below) and the radiation pattern.
Polarization is also a factor, where the sending antenna should be the same orientation as the receiving antenna. At 2.4GHz indoors this is not critical because we expect a considerable amount of polarization scattering due to multipath reflections. Also, the access point antennas are usually very flexible on how they are oriented which compensates for the lack of flexibility in the client device (ie, your speaker). The right angle antenna is actually polarized ok.
However, this antenna is presenting an unknown impedance due to the very close proximity of the reflecting metal keg, which appears to be about a quarter wavelength away, and also due to the long distance between the base of the antenna (the black part) and the ground plane (the keg). We antenna folks fuss about impedance a lot because it is one of the key things that contributes to how effective an antenna is. We seek "maximum power transfer" in both receiving and transmitting directions. Matching impedances from source (the transceiver), through transmission line (the coax) to load (the antenna) ensures that power is not wasted, it gives us maximum power transfer and minimum loss of power. Impedance matching is important because mis-matching costs you power, and power directly leads to range, and we all like a lot of range.
The coax cable you use has a characteristic impedance of 50 ohms, and you need your antenna to also have an impedance of 50 ohms, which is what I mean by impedance matching. In order for your monopole antenna to have an impedance close to 50 ohms, you have to provide it with a ground plane that is in the correct position and orientation. In this case you have not. As a result, you may be wasting anywhere from 0 to 10 dB. Its hard to say exactly how much.
Now, you could say that I'm obsessing over details and that the way you have it now should work adequately well, even if it is not optimal. That's true. But from a looks point of view, I think this is only a partial improvement over what you had before. You can stick with this if it looks ok to you, but it could be better. Its hard for me to argue for changes if you are happy with it.
I also want to mention that, if you are satisfied with the range, perhaps doing the change to how the coax connects to the module is not necessary. While doing the mod would optimize the connection (and possibly buy us an additional 1 to 10 dB of additional range), it also has some risk that we may break something that becomes hard to repair. Is the risk worth it? Then again, I think you mentioned that you were thinking of changing out to a new module anyways.