Rainbow, to answer your question generally, I meant that in any experiment it seems to me a good idea to be careful about measuring variables, maybe setting goals and tracking progress. A "good" cup of coffee can be "good" in many different ways, such as taste or health effects or social or cultural acceptance. Your question as to which variables is quite appropriate. When I read a medical study about a drug or intervention I look beyond the abstract to try to understand the dose used, for example.
I can only report what I track, and I record many daily "vital signs" such as blood pressure, body temperature, weight, waist, heart rate variability, exercise, etc., but my main interest in coffee/caffeine at present stems from about five years ago (before knowing genomic data) when I was concerned about developing Parkinson's Disease, because of episodes of REM sleep behavior disorder. There have been many studies about possible preventive effect of coffee/caffeine on neurodegenerative diseases (see paper above) and I enthusiastically embraced them as something I could control. The variables I tracked were related to sleep, such as sleep onset latency, daytime sleepiness and alertness, sleep length, deep sleep, REM sleep, apnea/hypoxia index (I use CPAP), weight, blood pressure and heart rate, and mood. I found these variables gradually improved with 5-6 cups coffee, intermittent fasting, low carbohydrate diet, less saturated fat, calorie restriction, weight loss to waist < 0.5 height, and changes in circadian rhythm (3mg melatonin when awaken in middle of sleep, not at beginning of sleep--I am genetically advance circadian rhythm, short sleeper--no daytime naps, 5 cups of coffee in morning and not after noon--I am fast coffee metabolizer--and delaying sleep cycle with blue light in evening not morning). For sleep, the Epworth Sleepiness Scale is useful:
http://epworthsleepinessscale.com/about-the-ess/
I also found I was having more GI symptoms such as urgency, diarrhea, and cramping when using 70 grams coffee a day and they went away when I reduced that to 30g, without affecting other variables such as sleep. I don't know the status of my microbiome otherwise.
Coffee variables have been studied by the Coffee Brewing Institute, for average consumers, and can be controlled to produce a perfect cup for your personal taste that also meets health needs. Several modern coffee makers such as the Behmor I use allow you to tailor the brewing temperature, time, grind, and variety, and you can learn the elements of taste such as acidity, flavors, robustness, etc., by experiments yourself, without having to rely on the Starbucks barista. As I said, I was able to experiment by measuring variables and seeing the results over a period of time, and I can make adjustments if necessary. Since my main goal is prevention, I really won't know myself until too late, of course, but meanwhile there are risk factors to control.
I was so anxious about my REM sleep behavior problems because of the risk of Lewy Body dementia or Parkinson's, and so maybe overdid the coffee drinking, but knowing I don't have genetic risk for those particular conditions now, and their disappearance with CPAP, fat loss, and other changes, has allowed me to feel more in control. But now I am anxious about late-onset Alzheimer's (e3/e4), and there are not so many biomarkers to be confident of control. (I have had a lot of tests from a clinical trial but await decision to accept).