The "logic" of hypotonic fluids is based on the Holliday-Segar methods of calculation for the daily needs of water and solutes based off normal, healthy adults and children. That being said, by definition of being hospitalized, a patient is not healthy and people argue (rightly so), that one shouldn't extrapolate healthy calculations into sick patients.
The risk of hypotonic fluids is the development of hyponatremia. This is a well documented phenomenon. This is likely a combination of the fluid themselves and the patients pathology. Considering respiratory symptoms (pneumonia, brochiolitis, asthma) are the most common causes of pediatric hospitalizations, and that lung pathology causes ADH release (through unclear mechanism) and can cause patients to have SIADH, the use of hypotonic fluids does have an increase risk of hyponatremia and hyponatremic seizures (I've seen it several times before).
However, patients who are ill, especially those with vomiting/diarrhea/increased urinary loss/insensible loss have free water deficits for which isotonic fluids may be inappropriate. Additionally, you have to remember that NS also has 154 mEq/L of Cl- which can induce a non-gap metabolic acidemia (though it is typically harmless). So the use of isotonic versus hypotonic fluids really depends on the patient and the pathology. Just like most of medicine, one size doesn't fit all.
I'm a big fan of Plasmalyte or LR. However, NS and versions of it are simply easier and cheaper to make, can be customized, and are most compatible with other drugs (due to lack of Ca++ and Mg++), and thus are more commonly used. I personally think it is okay to use 1/2 NS (but not 1/4 NS unless to are specifically correcting a free water deficit for a finite period of time... like diabetes insipidus) for short durations, but if one does so, they need to follow Na++ levels.
