This is just what i picked up from my colleagues at work while i was helping them with electronics trouble in their lab. They are doing research using attosecond length laser pulses, which is why they needed to use hollow core stuff to keep the signal fast and clean. If you dont remember what atto means those pulses are 0.000000000000000001 seconds long (well in that range not actually 1 attosec more like tens to hundreds of attosecs). But yeah thats all i know ^^

For more technical details from Wikipedia:

Optical fiber > Refractive index

The refractive index is a way of measuring the speed of light in a material. Light travels fastest in a vacuum, such as in outer space. The speed of light in vacuum is about 300,000 kilometers (186,000 miles) per second. The refractive index of a medium is calculated by dividing the speed of light in vacuum by the speed of light in that medium. The refractive index of vacuum is therefore 1, by definition. A typical single-mode fiber used for telecommunications has a cladding made of pure silica, with an index of 1.444 at 1500 nm, and a core of doped silica with an index around 1.4475.[65] The larger the index of refraction, the slower light travels in that medium. From this information, a simple rule of thumb is that a signal using optical fiber for communication will travel at around 200,000 kilometers per second. Thus a phone call carried by fiber between Sydney and New York, a 16,000-kilometer distance, means that there is a minimum delay of 80 milliseconds (about 1/12 of a second) between when one caller speaks and the other hears.[c]

Thanks for the source :)

Starlink’s LEO satellites will be in 1,110 to 1,325 km orbits. Although much lower than GEO, this is still too high to pro- vide lower latency than fiber over shorter distances. However, over longer distances the extra latency getting between Earth and the nearest satellite may be more than offset by routing around the world between the satellites at c.

So yeah makes sense. The inital distance to space completely tanks the short distance latency, but eventually gets compensated by the higher transmissions speed when trying to cover longer distances. I havent found it yet, but what i really wanna know is the distance value at which sat overtakes fiber. I dont think this paper covers that sadly.