Precisely this. I have done several experiments over the years and can really attest to that - however I learned very early that saying that in public is begging to be laughed at due to the common misconception. The tap water in my home town was a very poor conductor and even the hard water I have here in Dallas is better, but nothing to call home about. That, tied to the much lower resistance path between the resistive element and the ground or between the live and neutral, guarantee that millions are not zapped every day by these chuveiros.
Yes, this is certainly true if the device is operating as intended. There is a problem, though, if the resistance heaters were to break off or corrode: the neutral side of the element might be covered in scale and become a poor conductor, or be corroded away by impurities. If the ground wire is left disconnected (which seems common), or it too is corroded, then the only path from live to ground is the user. Even if the ground wire is connected, it doesn't do much to shield the user: it simply pokes into the stream of water. This would be mitigated almost entirely by the use of a sensitive RCD (5mA trigger), but without that it seems too hazardous.
Water quality is taken for granted, but it can be quite variable: as can its conductivity. We have no problems with our water here, but every year when the distribution system is flushed, the dissolved metals go up a lot and it turns brownish. What could seem electrically safe on some days might be dangerous on others.
If water was so good of a conductor, why would you die of electrocution if your body was completely wet? Wouldn't electricity flow mostly through the water drops in your skin instead of through your heart?
The body's contents inside the skin is a very concentrated aqueous solution of electrolytes. Ionic solutes increase water's conductivity a lot. The current that can be lethal to the heart is also very small, only a couple of mA directly. In no way does most of the current need to travel to the heart to cause death.
Believe me, water quality varies greatly in Brazil, so much that a common misconception there is that it is not potable (people usually don't drink from the tap, but instead put some "filters" that do nothing more than block residues).
What I discuss is how efficient a "path to ground" or "path to neutral" through one's body would be instead of much lower resistance paths. The galvanized iron connections on drains haven't been used in decades now, as well as the fact that almost 100% of shower stalls are porcelain or granite, which increases the resistance to ground by an enormous extent.
I agree the body and the inner skin are very good electricity conductors, but the outer skin... Not so much. Again, which one would be a lower resistance path?
Look, I think we all understand the common
chuveiros are more dangerous than a properly designed water heater, especially considering the quality of the installations around. However, at a certain point in time we must ask ourselves how a red herring this becomes, especially when facing the statistics. Alternatives also have their share of problems: in the past gas was widely used in water heaters, but they are inherently volatile and poisonous (my dad almost died of poisoning while taking a shower). Solar is initially expensive and subject to elements, which increase its cost of ownership (although some folks are reporting the electricity is becoming more expensive, favouring this solution). There is no perfect solution.
(edit: removed a silly sentence constructed)