Just a quick thought: the decision whether to produce or abort this product would be contingent on two notions:
1. The capacity to flexibly react to demand fluctuations (i.e., close to zero set-up cost for production capacity)
2. The possible speed of production (i.e., can an observed demand peak be almost instantly catered for?).
If these two basic conditions are met, and if the profitability of the product is attractive, then keeping this product in the portfolio might make good sense.
Hope this helps!