If you don't disguse the traffic (Spoof/Stealth) then going through a node buys you very little because the tracker can just watch for an easy to spot correlation between stuff going in and stuff going out.

See that as you may, that is how the rules currently work.

Actually the rules in the core book are such that it doesn't matter if you go through 50 hops (login or not) or none. It comes down to the Track Extended Test, and the only thing that makes that more difficult is the Stealth of the target (which subtracts dice from the tracker's pool) and the target scoring net hits on a Redirect (Hacking+Spoof vs. System+Track) action which increase the theshold of the Track Extended Test.

Track chews up and spits out hops like they weren't even there. No login checks required.

What my suggestions are doing is actually simulating benefits of those hops, since the core rules just don't have them (and Unwired is minimum a year away). Roughly speaking the rules are:
1. A hop does you very little unless it is has a very large volume of traffic, and has sizable processing capabilities to allow Stealth and Spoof to execute their funky stuff on the node to disguse the traffic. Rule of thumb is the processing capability to run the Stealth and Spoof programs you are using (but without the bother of checking the node for program load vs. System).
2. The more background traffic between the hop and the target system the better (this ties in with the "trusted" angle, and also using the traffic as cover)
3. A system has an inherent security paranoia rating that is not expressed in terms numbers, but instead just making it general knowledge that a semi-experienced decker would know (general trade knowledge) that it is nigh impossible to pull off an illegal entry without specified benefits using rules #1 and #2.

Currently i don't assign any dice benefits, i just lay out the narative for the decker about what he figures out to make 1+2=3. Another decker with a better Hacking skill might figure out a superior set of hops, but that's all factored in by the dice each decker rolls for a Redirect (if they actively take a Redirect action to take advantage of this "superior hop planning").

EDIT: Why no extra dice benefits for #1 & #2 beyond what #3 "requires"? Because creating numbers would tend to players wanting to me to tell them those numbers, which is certainly fair. That would require solid fleshed out numbers to stand by that bring with them the risk of not only skewing dice pools, but in the end just creating a set of standard hop actions that get used every time. Basically making number crunching busy work for the player, or effectively penalizing him for not taking the time for number crunching busy work.

EDIT: Likewise for why i haven't bothered for the dice pool modifiers for #3. I'll give a player an option if they ask for it, i'd do it if they actually wanted to consider going against their the advice of the "little voice" in the character's head. Or professional gut instinct if you want to call it that. Or in OOC terms GM definition of the world. However i'd make clear that it was case by case, and would not be ready to give hard and fast numbers. I'm a bit more comfortable with not giving penalty numbers, especially when i always provide them with the dice pool modifier neutral solution so they don't have to guess or just repeat the same thing over and over each senario.

EDIT2: Once the decker's player become comfortable with it i suppose he could be helping with the narative with suggestions of 1 & 2 to help overcome 3. Who comes up with it is really about the GM/playing style of the table and how experience the player is.