Yes, pancreatic beta cells can indeed sense the level of blood glucose. They respond to high glucose concentrations by secreting insulin, and stop (or reduce) secreting insulin when glucose concentration falls. This is a classic example of feedback regulation (homeostasis) in physiology.
Exactly how the glucose sensing in beta cells has been the subject of much research for decades; a freely available review is here. You are correct that phosphorylation and further metabolism of glucose by the beta cells is required for sensing. The current theory is that mitochondrial oxidation of glucose leads to increased ATP levels, which is detected by ATP-sensitive potassium channels at the plasma membrane, which in turn leads to secretion of insulin.
When glucose levels fall, beta cells reduce their glucose metabolism because they use the glucokinase enzyme to phosphorylate glucose, and glucokinase is a low-affinity enzyme that becomes inactive at lower glucose levels. Therefore, uptake and oxidation of glucose is reduced, ATP levels fall, and insulin secretion stops.
This is the current consensus I believe. But these metabolic processes are difficult to study, and the precise nature of the connection between glucose metabolism and the ATP-sensitive channels is still somewhat controversial.