I guess I'm wondering, if the mechanical limit of my mount is ~0.5" RMS and that can be acheived with a 130mm FL guidescope, what is the benefit from using a longer FL?
Although with a longer FL I wonder if 0.5" is the limit? Not that I need that accuracy. I've gone down as low as 0.3" to 0.4" on exceptional nights.
I suppose there must be a limit on the subpixel resolution the guiding software can achieve so if the mount is up to the task then a longer focal length would be necessary.
The rule of thumb is all I've ever seen in determining how effective a guider will be, and that's only dependent on the relative pixel scales of the imager and guider. Partly it's saying that if your imaging scale is "x", you need better than 5*x (or 3x or whatever) to keep your stars sharp. But I suppose on the other side, it could also be saying you don't need to be any better than that, so save your money? The OAG fans would probably disagree...
So, an interesting question. Let's think about this. What does the manufacturer mean by the "mechanical limit" of the mount? A single number is practically meaningless unless one knows the conditions under which it is measured. As you say, the mount and the particulars of the night both contribute to it. Probably more so, what you have positioned on top of the mount, in terms of weight and balance. All those things only seem to make things worse, so I'm guessing they're talking about the best case? I searched, and can't find such a spec for my AVX. Probably not a surprise.
Then I'm curious how one can achieve any particular RMS with any particular guidescope and camera. If indeed one is accurately measuring the sky to sufficient accuracy, sure, there's probably no benefit in measuring it more accurately. But a 130mm focal length and a camera with, say, 3.75u pixels, has a pixel scale of 5.94 arc-seconds. You'd be pushing the software to measure down to 0.5". That said, my guider arrangement has a pixel scale of 4.3", and according to PHD2 I'm getting an RMS of 0.5" in Dec and 1.5" in RA, but I have no idea how one translates (predicts) one set of numbers to the other. There's a lot of mechanical stuff in between.
I just go with the rule of thumb and then optimize the stuff around it - balance, compact arrangement of accessories, etc.