Let me first state that I have no idea really what goes into the development of a product like Survey123, which means that this suggestion may be totally unworkable, but here we go:
Allow "operational rows"
Many of my surveys are build on referenced feature services, which have a pre-existing data schema including the data that we actually need for that service. In order to use Survey123 to its greatest potential, it's often necessary to add rows (corresponding to attribute columns) to accomplish things like calculations or to hold a value that gets used by a different question in the survey. These rows might be totally useless other than to provide context for a different question.
As an example (uncommon scenario, but play along), say I want to get the number of the current day of the month so I can use that in a calculation, say to concatenate a unique ticket number. In order to do that, I need a row that gets a date, either the current one or one I select. If I don't actually need that date field for anything else, it just sits there, a requirement of a good survey form, but important to my dataset overall. Or perhaps I am using an API to get real time data - I will need a row to hold a JSON response, but the weighty, long text field is otherwise unimportant after the fact.
What I would love to see is the ability to add lines to a Survey that DONT correspond to a data column in the feature service that can be used to hold values, run intermediate calculation, etc, so my dataset isn't cluttered by Survey123-specific needs. We take great care to build good datasets, and Survey123 should be a way to extend them, not to bloat them for the purpose of a usable survey form.
