I would approach the types of cases differently. The first one you mention I would call Diagnosis cases, which require you to diagnose the root cause of a problem and offer a solution. These may be related to profits, or not (i.e., why is our client experiencing delays in its production process?).
The other type of case is related to a Conceptual framework, which typically tackles open questions. Some examples: should we enter a new market? Should we launch this product? How would you improve the educational system of your country? etc. For these cases, there is not one single answer and it typically does not involve quantitative components and is more qualitative and conceptual in nature.
For BOTH these cases, hypotheses are crucial!
For Diagnoses cases, the first step consists in understanding, NUMERICALLY, where the problem is coming from. Profits are down? How much is due to costs? How much is due to revenue? In Revenue, how much is due to price fluctuation, and how much is due to demand fluctuation? etc. This first part involves no hypotheses in itself, you just want to find out what the data is. The hypotheses come when you have to interpret the data. Perhaps sales are going down because there is a new competitor in the market? Perhaps our client has suffered a PR problem and clients don't want to buy our product anymore? The hypotheses are endless, and to develop the right one you need to have business acumen and understand what the data is pointing at. The last part of these types of cases revolves around the solution to the diagnosis, and these will depend on both the data and the hypotheses you found to explain the data.
For Conceptual problems, hypotheses are relevant in EVERY PART of your framework, otherwise, there would be no reason to mention them at all. I will give you an example:
Should Ferrari launch an SUV?
A basic framework could cover MARKET, COMPETITION, COMPANY, CLIENTS.
The Hypothesis behind these buckets is the following:
MARKET: If the market is attractive, then Ferrari should launch an SUV. So your questions in this bucket will test this hypothesis.
COMPETITION: If Ferrari can compete in this market AND the market is attractive, then it should launch an SUV.
COMPANY: If the market is attractive AND Ferrari can compete AND the company has the necessary competencies, know-how and resources to successfully market an SUV, then it should launch the SUV.
CLIENTS: All of the above PLUS If our clients would value a Ferrari SUV and this would not negatively impact other sales, Then it should launch the SUV.
These hypotheses are crucial in determining what questions I will ask the interviewer in each of these buckets, and serve as the overall guideline to your reasoning and communication.
If you don't have a hypothesis behind a conceptual question you are asking, then, logically speaking, your question is irrelevant.