Well enough is not universally defined, though. One person may just want something to mask ordinary human odor. Another might have allergies, or has friends who have allergies. Various people respond differently to different fragrances. Some fragrances might work better in a professional vs. dating environment. Etc. Objective data could resolve all of those questions far better than any sex-driven superbowl ad.
In the case of deodorant, the question one really wants to ask is either "Will this fragrance upset my coworkers with allergies," or "What fragrance will most impress the guy/girl I want to impress?" Objective, empirical data from the nearest demographic comes the closest to answering that question.
There's no reason that objective data has to require time spent. Computer analysis and machine learning could present the right data at the right times, if the algorithms were designed for that instead of for behavioral manipulation. You wouldn't go to the store, see a rack of 150 different varieties of deodorant, and read the data for each one. You would specify your requirements and see a short list of options, with the principal component of the remaining differences emphasized in the displayed statistics.
More effort is put into seductive advertising than would be required to build and present empirical analyses of fundamental product attributes, despite the perception of complexity.
To address your final paragraph, products that offer no meaningful distinction from their competition would die, while those that are truly useful (even for highly subjective definitions of "useful", like "fashionable") would thrive.