I think you may be missing the point (no pun intended) of fixed point. The point is an imaginary concept that you carry through your computations. Your your case, you literally have to do nothing.
For your VHDL, if you are using a common format through out the pipeline, the only scaling you may need is after multiplies. (And even then you may retain precision in any accumulation, etc you may do).
Now, technical speaking when you say Q2.10, that is really a 13-bit number. The Q format generally implies you are working with signed number.
If you are filtering, you don't have to even work about your 0.0005v scaling to volts. That is somewhat meaningless for the purposes of the filter. All you have to worry about is how you scale factor may change in your filter pipeline.
Here is a point to ponder. What is the difference between a Q2.10, Q3.9 and Q4.8 number? Absolutely nothing. They are all 13-bit integers. It is purely how we chose to interpret them.