Is tolerance not an issue with those? I have heard of the wonderful effects on sleep that gabapentin has (that alone entices me) but I am not totally convinced that GABA based medicines don't come with rebound issues. Even Picamilon can have that effect for me sometimes and that one is fairly benign in comparison.I may have missed this reading the forum but have you considered bipolar type II. Agitation and anxiety on SSRIs and in your case more then one is usually a sign of a mood disorder. Have you been on a mood stabalizer such as lithium, valproate, carbamezapine, topomariate, gabapentin (controversial), pregablin (conterversial), lamotrigine, abilify?
I might give picamilon or phenibut a go for a few days and see how you feel. If you start to cheer up on these I would reccomend getting on something GABAergic.
The broader issue is not exclusively about GABA, but the neural excitation/inhibition balance. Gabapentin and Pregablin are not simply GABAergic, but are also Ca++ channel blockers, which gives them some of the properties of NMDA antagonists like Memantine. On the minerals front, Mg++ and Zn++ are also Ca++ channel blockers, but it is controversial whether supplements merely correct deficiencies or can flood the brain like drugs.
Similarly, mood stabilisers like Valproate are Na+ channel blockers, with the corresponding minerals analogue being Li+. (This involves loads of simplifications, but you get the idea). Li+ also seems to act as a NMDA antagonist for reasons that escape me, and I'm not even sure whether it is regarded as an essential trace element.
Another way of skinning the inhibition cat is to antagonise noradrenaline. Beta blockers work centrally and peripherally, whereas agents like Clonidine work centrally only. There are tolerance and rebound issues with chronic usage here.
Lastly, 5-HT1A agonism can be anxiolytic, with Buspirone being the most specific. The nice part is that tolerance doesn't seem to be a big problem, but the downside is that it seems to be tricky to get it to work properly.
The main point I'm trying to make is that your options are not limited to habit-forming tranqs.
The problem with Buspirone is that it's a rather weak agent, and although it may appear to have some novel appeal as an adjunct, I think most of the more common serotonergics should do the trick. And it should be noted that magnesium has much more of a modulatory effect on calcium channels. Finally, the binding affinity of these drugs should be more closely examined, because altering levels of glutamate in some regions should not be conflated with a strongly antagonistic affinity for the NMDA group of receptors.