You could ground everything and still have an earth loop, earth loops are caused by different length (or more accuratly different resistance) earth cables from each piece of equipment. That is why studio power should always be star, from a single scource / power point. That ensures ahat as close as possable the earth is the same length.
Absolutely true. 12-14 AWG stranded or solid copper wire at pretty much any practical length tends to have significantly lower resistance than braided or foil cable shields between devices in most environments, which is why even very sloppy star grounding works 99% of the time. The exceptions can cause much pain and suffering, though.
Unfortonatly some earth cables have crap thin wires (like internal computer earthing) and some have big fat copper cable (like pro gear) so the earthing resistance is different.
If your audio interface has its own ground (as it should), this shouldn't matter. Any digital bus noise should get earthed long before it can impact the audio. Actually, this shouldn't matter if the audio interface is designed correctly anyway, but many (most?) aren't, so....
Early laptops had double insulated power supplies with no mains earth, never had a earth loop problem, the new ones are mains earth, with the mains earth pin joined to the negative on the low volt side of the power supply, and now there are all kinds of earth problems with the biggest being digital noise from USB sound cards due to an earth loop via the USB cable.
The main reason that some older laptops had fewer problems is that newer laptops incorporate CPU and GPU clock stepping, voltage stepping, powering cores up and down,, changing bus speeds (and thus RAM power utilization), and other power management techniques, many of which generate
huge spikes on both the power and ground busses. Also, newer operating system versions and drivers take advantage of these technologies to a greater degree. The result of this is that newer laptops produce orders of magnitude more digital noise than you could possibly generate with an older laptop even if you tried.
This is further compounded by a push to design cheaper and cheaper power supply circuits to undercut the competition, resulting in less noise damping on the digital grounds than you might have seen in previous generations of hardware.
Top that off with the desire to squeeze more functionality out of USB and FireWire bus-powered devices. The more current the device can potentially drain during operation, the more trouble you're going to have with inrush current exceeding the USB or FireWire specifications. To get around that, the usual fix is to reduce the amount of power filtering. When you do that, you get more digital bleed. This is yet another reason why bus-powered USB and FireWire devices are always something of a compromise.
In my experience, digital noise problems are usually worse if you disconnect the ground on a three-prong modern laptop, though I suspect you could come up with an exception if you tried.
When you refer to cheap ungrounded gear don't be too general, I have pro gear with no mains earth, the reason is that they have a double insulated internal power supply, a internal floating ground (nothing uses a chassis ground as a signal or power return) and therefore has no source for an earth loop.
Bear in mind that the term ground "loop" merely refers to a ground between two pieces of equipment with a high resistance to true ground allowing signal to flow from one device to another across what should be a ground. It does not require a true earth on each device, nor does it require an actual loop, per se. By definition, the audio cable between any piece of gear with a floating ground and any other piece of gear
is a "ground loop" by itself; any noise introduced by the device with the floating ground will flow away from the signal ground in the next device over.
Just to be clear, I'm not saying it's not possible to design a two-prong device that rarely causes problems. I'm just saying that in every case where I've actually encountered ground hum, it involved two-prong or wall-wart-powered gear (some of which was, in fact, supposedly "pro" gear).
