1: \begin{abstract}
The inability of primordial baryonic density fluctuations, as observed in the cosmic microwave background (CMB), to grow
into the present day astronomical structures is well established, under Newtonian and Einsteinian gravity. It is hence customary
to assume the existence of an underlying dark matter component with density fluctuations, $\Delta(M)$, having amplitudes much larger 
than what CMB observations imply for the baryons. This is in fact one of the recurrent arguments used in support of the dark
matter hypothesis. In this letter we prove that the same extended theory of gravity which has been recently shown to accurately
reproduce gravitational lensing observations, in absence of any dark matter, and which in the low velocity regime converges
to a MONDian force law, implies a sufficiently amplified self-gravity to allow purely baryonic fluctuations with amplitudes in 
accordance with CMB constraints to naturally grow into the $z=0$ astrophysical structures detected. The linear structure formation 
scenario which emerges closely resembles the standard concordance cosmology one, as abundantly calibrated over the last decade to 
match multiple observational constraints at various redshifts. However, in contrast with what occurs in the concordance cosmology,
this follows not from a critical dependence on initial conditions and the fine tuning of model parameters, but from the rapid 
convergence of highly arbitrary initial conditions onto a well defined $\Delta(M,z)$ attractor solution.
\end{abstract}