Since $x<0$, let $x=-y$ and use composition of Taylor series around $y=0$
$$2^{-y}=e^{-y \log(2)}=1-y \log (2)+\frac{1}{2} y^2 \log ^2(2)+O\left(y^3\right)$$
$$1-2^{-y}=y \log (2)-\frac{1}{2} y^2 \log ^2(2)+O\left(y^3\right)=y\log(2)\left( 1-\frac{1}{2} y \log (2)+O\left(y^2\right)\right)$$
$$\log(1-2^{-y})=\log(y)+\log(\log(2))+\log\left( 1-\frac{1}{2} y \log (2)+O\left(y^2\right)\right)$$
$$\log(1-2^{-y})=\log(y)+\log(\log(2))-\frac{1}{2} y \log (2)+O\left(y^2\right)$$
$$\log_2(1-2^{-y})=\frac{\log \left(1-2^{-y}\right)}{\log (2)}=\frac{\log (\log (2))}{\log (2)}+\frac{\log (y)}{\log (2)}-\frac{y}{2}+O\left(y^2\right)$$ which seems to be quite good
$$\left(
\begin{array}{ccc}
y & \text{exact} & \text{approximation} \\
0.05 & -4.87562 & -4.87569 \\
0.10 & -3.90041 & -3.90069 \\
0.15 & -3.34008 & -3.34073 \\
0.20 & -2.94954 & -2.95069 \\
0.25 & -2.65196 & -2.65377 \\
0.30 & -2.41313 & -2.41573 \\
0.35 & -2.21480 & -2.21834 \\
0.40 & -2.04608 & -2.05069 \\
0.45 & -1.89993 & -1.90577 \\
0.50 & -1.77155 & -1.77877 \\
0.55 & -1.65754 & -1.66626 \\
0.60 & -1.55535 & -1.56573 \\
0.65 & -1.46307 & -1.47525 \\
0.70 & -1.37922 & -1.39334 \\
0.75 & -1.30259 & -1.3188 \\
0.80 & -1.23226 & -1.25069 \\
0.85 & -1.16743 & -1.18823 \\
0.90 & -1.10745 & -1.13077 \\
0.95 & -1.05180 & -1.07777
\end{array}
\right)$$
If more accuracy is required, you could continue the process and get
$$\log_2(1-2^{-y})=\frac{\log (\log (2))}{\log (2)}+\frac{\log (y)}{\log (2)}-\frac{y}{2}+\frac{1}{24} y^2 \log
(2)+O\left(y^4\right)$$
For illustartion purposes
$$\int_0^1 \log_2(1-2^{-y})\,dy=-\frac{1}{2}-\frac{\pi ^2}{12 \log ^2(2)}\approx -2.21186$$ while the last approximation to $O\left(y^4\right)$ would give
$$-\frac{1}{4}+\frac{\log (2)}{72}+\frac{\log (\log (2))-1}{\log (2)}\approx -2.21183$$