- whether offering it online significantly facilitate students attending;
- whether online contents enriched or elevated the core of the course.
A course could go online if it passes one of the Ort Test. However, if it passes neither, the school should not just go ahead because of monetary concerns as that may very well backfire in a not so long run.
Other thoughts: when a professor teaches, more or less he brings in the knowledge and methodology he absorbed from his PhD adviser. Assuming only a limited number of schools (300 or so in the US) produce PhD, we can count on teaching efficiency of any PhD recipient. However, when the entire higher education is rushing to online offerings, few current professors have been trained or even had any level of experiences of online offerings. Many are retooling from whoever have been conducting online education. Unfortunately, the majority of the said 'whoever' are for-profit institutions and community colleges, or bottom level schools otherwise. Following models of such bottom level schools will certainly not help maintaining quality or reputation. Instead, an efficient and cost effective way of training is giving each faculty who are interested in offering online courses a chance to take one at a first-class school, such as Stanford or MIT. If that's not possible, inviting professors from those schools to give a talk may also help.