What I'd like to clarify when we are talking about the efficacy of the sat trackers is - are we talking about one installed on the aircraft with an external aerial or are we talking about a personal one stuffed into a pocket?
I've been talking about the pocket-size variety, but have briefly mentioned more complex equipment, the most typical example perhaps being the Spidertracks brand.
Let me also preface my response by I am assuming that a pocket size unit is not buried in some ridiculous place, but perhaps in a shirt or sleeve pocket.
I should imagine the signal is somewhat different and therefore the tracking resolution varies.
This is a complex question that divides itself into three parts.
Part 1: GNSS positioning accuracy (I write GNSS because some of these units utilize more than just the GPS constellation. For instance many of the more expensive Garmin units use both GPS and Galileo).
GNSS positioning accuracy is affected by many factors. However, in this case we are only concerned with how many satellites the unit can see simultaneously. More is better because a) it can make more measurements, and b) it can choose more measurements with more favorable geometries. To use a very loose analogy that may be familiar to you, think more LOBs, and LOBs that have better spacing.
Certainly an external antenna will be able to have a direct view of more satellites than one in cabin. Nevertheless, because there are now so very many GNSS satellites in orbit, it makes very little difference in practice. We are talking maybe a 10 or 20M difference, at most, if even that much. Once on the ground, experience has shown that externally mounted antennas fare poorly in crashes, so there's another difference to consider.
Part 2: Iridium link resilience
Again, it's a similar answer to Part 1, but in this case it's a question of if any updates are delayed or missed. There are normally only 1 to 3 Iridium satellites in view at any given time. In addition, in the cases where there is only a single Iridium satellite in view it tends to be at a high angle in the sky. This definitely favors an external antenna. However, if the pocket sized device is in, say, the chest pocket of the pilot, and the helicopter is of a typical civilian nature with a large forward bubble windscreen, it's not that big of a difference.
Part 3: position update rate
With an inReach or similar product they only offer a 2 min. update rate, and that's an average because of "Part 2" effects. This is primarily driven by battery life, but also because more messages cost more. With something like Spidertracks, you can pay (quite a bit) extra for 30 second update rates. Again, that's an average because Iridium short message service completion times vary.
Of the three parts of the problem, this last part would make the biggest difference, of course. But this matters only if there is no active beacon or tracker signal emanating from the accident site.
Interestingly, I would be willing to bet a beer or two that most Spidertracks installations do not use an external antenna. These units tend to get installed with internal antennas in the area of the chin bubble or instrument panel in order to avoid the paperwork and expense of external aircraft modifications. We've got a transient 206 in the hangar like that right now. However, I could not tell you what sort of update rates most Spidertracks (or similar service) users typically pay for.