ACR doesn’t work that way. There is an eyedropper for the overall white-balance that you can set by clicking on a relatively bright, neutral area, but there aren’t black and white-point eyedroppers. But, what is clipped black or white is determined by the output colorspace you have set at the bottom and you can get some idea of what will be clipped black and white by turning on the clipping indicators and watching for the areas you’d be eye-droppering on (the black and white patches in your grayscale) to start showing the clipping color overlay for black and white, but any further toning adjustments will move data in and out of what is darker than black and lighter than white so what is clipped for output will change as you fiddle with the various toning sliders.
The reason things are different: In PS, when you set a black-point and white-point in Levels and click Ok, the data below and above those is thrown away and that is the starting point for the next adjustment. In ACR, things are non-destructive and all the adjustments occur simultaneously, rather than sequentially as in PS, and no clipping occurs until you create the output file. This means that the pixels that are darker than your black-point and lighter than your white-point are still there, and can be moved into the black-to-white range if you do further toning adjustments. Having more headroom of the pixels that are outside the black-to-white range is good because you won’t get the posterization or banding that occurs when colors and levels are clipped after each adjustment in PS.
Can you describe what you intend your overall workflow to be on an image: how much adjusting will occur in ACR and how much, if any, might happen in PS, after the image is opened into PS from ACR, or opened from a file saved from ACR? There might be other ways to approach what you’re trying to accomplish.