If anybody is still interested: this is all about the partial/apparent specific volume of NaCl in water.
Your first and main mistake is assuming that the $\pu{9 g}$ of NaCl are dissolved in $\pu{1000ml}$ of water. Since the total volume is $\pu{1000ml}$ and the NaCl contributes some volume, it‘s less water (see below). Your second main mistake is assuming that adding $\pu{9g}$ of NaCl doesn’t change the volume of the solution, and thirdly, you confuse SG (relative density) with density.
How much volume the NaCl contributes is determined by its partial specific volume. At infinite dilution, you might expect the partial specific volume to be the reciprocal of the density, but it is often less, because the dissolved solute can take up some of the space between the solvent molecules, or it can be greater if the solute molecules tend to repel the solvent molecules, or the ions take up more space then they do when combined. The partial molar volume is greatly determined by intra-molecular forces. Hence it is for all intents and purposes incalculable and an empirical quantity.
Let‘s do the calculation:
You say the solution has an SG (relative density, i.e. relative to the density of water) of $1.0046$, but a cursory search leading to tables of density of NaCl solutions suggests this value is the actual density of the solution at 22 °C (I found a table of density which is cited as having been taken from "Perry's Chemical Engineers' Handbook" by Robert H. Perry, Don Green, Sixth Edition). If so, one liter has a mass of $\pu{1004.6g}$, and hence contains $\pu{1004.6 g} -\pu{9 g} = \pu{995.6g}$ of water.
The density of water at 22 °C is $\pu{0.99777 g ml-1}$, so this water has a volume of $\pu{995.6 g}/\pu{0.99777 g ml-1} = \pu{997.8ml}$.
Since the total volume is $\pu{1l}$, the $\pu{9g}$ of NaCl in the solution have resulted in a volume increase of $\pu{1000 ml} - \pu{997.8 ml} = \pu{2.2ml}$, hence NaCl at this dilution has an apparent specific volume of
$$\frac{\pu{2.2ml}}{\pu{9g}} = \pu{0.244 ml//g}$$
This agrees well with a simple article I found from a student („Solution Density and Partial Molar Volume as Functions of Concentration“ by Melissa Michaels) in which the author uses standard methods to determine the partial molar volume of NaCl at various concentrations based on experimental measurements of the density. NaCl in the solution with $\pu{10.34g}$ NaCl in $\pu{1000ml}$ of solution was shown to have a partial molar volume of about $\pu{13.68 ml mol-1}$ which would be a partial specific volume of $\pu{0.234 ml g-1}$.
The density of NaCl is $\pu{2.16 g ml-1}$ and the reciprocal of this is $\pu{0.4629 ml g-1}$, so the ions seem to be fitting in between the water molecules well and/or attracting them, which makes sense since NaCl is an ionic compound and water is a polar molecule.
These calculations and this statement are based upon the assumption that the true density of the solution at 22 °C is $\pu{1.0046 g ml-1}$.