So, before I get into the specific stat priorities and the whys and the what-to-dos, I want to address a specific misconception in your question:
For instance, I don't need to hit the 'hit cap' for poisons (spells) so I'm suspicious of the purported importance of the expertise cap.
Getting to the hit cap is just as important for dealing maximum damage in normal 5 mans, as it is in heroic 5 mans, as it is in raids. What differs is not whether you need to get to the hit cap, but what the hit cap is. More specifically, since you're not raiding, you probably don't want to worry about being spell hit capped for level 88 boss mobs, which is what the theorycrafting at forum like EJ is going to be centered around. You most certainly do want to be capped for level 87 mobs, which is what most Heroic bosses are.* Fortunately for you, this cap is much lower - only 6% spell hit and 6% melee hit, or 241 hit rating with the standard 2/3 in precision.
*Heroic end bosses are in fact level 88, much like raid bosses, but it's probably not worth optimizing for them, and thus, wasting a truly staggering amount of itemization on hit.
Now, as to the specific question of whether expertise matters: the short answer is NO. Because the majority of the cost of dodged and parried special attacks is refunded, the overall DPS cost of not being expertise capped is relatively low. By this measure, Expertise is among the weakest DPS stats, ahead of only Hit above the spell hit cap and Crit. That said, there are several specific reasons that you might choose to ignore this advice and cap expertise anyway.
The first is cycle stability. While it's true that the DPS cost of a dodged Envenom or Mutilate is relatively low, the fact that it is possible for these abilities to miss requires you to pay additional attention to managing your rotation. This additional attention is attention that you may not be paying to, for instance, not standing in the fire. Thus, many players do choose to place increased value on Expertise to minimize the amount they need to watch timers and combo points. Personally, I tend not to place much weight on this argument, and I only include it here in the name of completeness, but it isn't entirely without merit. I'm simply of the opinion that one is better served learning to manage their rotation with the randomness while also paying attention to their surroundings.
The second reason you might find is more commonly found at the raid level - if you are working on content during which you are spending a disproportionate amount of time attacking from the front - Aragloth, Chimaeron, Maloriak, or Al'Akir depending on your strategy for instance - the value of Expertise rises somewhat - that said, this increased value is not so massive as to substantially rearrange stat priorities. At best, it makes Expertise more desirable than haste, and perhaps strengthens the arguments that can be made regarding cycle stability. For Heroics however, this should be a non issue.
The final, and in my mind, most legitimate reason to consider Expertise is actually specific to a rogue that is spending a substantial amount of time in heroics: Shiv. There are a number of Heroic encounters where you may be called upon to Shiv an enraged boss or mob, in order to protect your tank from being summarily splattered. If you truly find dispelling these in a timely manner to be an issue, adding more expertise to allow Shiv to apply more reliably may help. Again, I can't recommend this, but I'd acknowledge it as being a useful reason to not neglect Expertise entirely.