An "underpenetrating" round that only goes in 8" into the body cavity is entirely sufficient … .
A bullet that penetrates 8 in from a frontal shot does not underpenetrate. Eight inches of penetration into the chest from the front is plenty to reach vital tissues of all but the most massive land whales (who likely don't have the mobility to assault you anyway).
What is desireable is a round that will penetrate adequately from a suboptimal angle, e.g., from the side. If a reasonably large person is 24-in wide from outer arm to outer arm, 12 in penetration would be nice to reach the center of mass of vital tissues within the chest cavity, but something less than half way through will still reach one lung.
Don't confuse gel penetration with actual penetration into a body. As the professor who founded the Statistics Department at my alma mater was fond of pointing out: All models are wrong (after all, they are simplifications of reality), but some models are useful. Standardized gel penetration testing is useful in selecting effective handgun rounds for wounding. The International Wound Ballistics Association — comprised of experts in the field of human terminal ballistics — observed that those handgun round that produced at least 12.5 in of penetration in standard, calibrated bare gel, and at least 13.0 in of penetration in such gel preceeded by four layers of heavy denim, were capable of penetrating deeply enough into a human to cause effective wounding.
I don't buy the FBI's undocumented modification of the IWBA criteria — to 12.0 to 18.0 in — because I don't know their rationale. I'll stick with the documented opinion of the experts. I believe the FBI's upper penetration criterion arises from the assumption that more penetration than that in gel will likely pass through a human. In comparing permanent wound channel volumes between rounds — depth of penetration in inches is multiplied by cross-sectional area of the expanded bullet in in^2 — I've seen the FBI cap the depth penetration at 18.0 in for bullets penetrating further.
I don't care about the FBI's barrier penetration testing. If I'm concerned about penetrating barriers I'll carry a magazine with FMJs, or candy stripe a mag.
And, I'm not concerned about overpenetration. Any bullet I fire that strikes my assaillant and passes through him will have little remaining energy with which to wound another. And, I wouldn't fire if an innocent bystander were in the target background. Besides, even the FBI reports that the hottest gunfighters they field miss their target 70% of the time with each trigger pull. Their poor gunfighters miss at a rate of 80%. When I began reading gun magazines in the early '80s, the average LEO gunfight miss rate was reported to be 83%. Each missed shot has all its energy. Thus, anyone willing to use a firearm in defense better wed himself to the rule of knowing he's not jeopardizing anyone downrange of his target before pulling the trigger.
All things considered, I'd carry FMJs in any .380 pocket pistol. Some very few brands can be shown to penetrate adequately while expanding reliably, but often various testers report different results. Tests are also highly dependent on barrel length. If your barrel is shorter than the barrel of the test pistol your rounds may not expand, in which case you are better off carrying cheaper FMJs. If your barrel is longer you may get greater expansion, to a degree where penetration is inadequate, in which case you are better off with FMJs, as penetration über alles in the handgun defense game.