From Mary, et al., 1996, Figure 1 on the second page shows a nice breakdown of there the Carbon and Nitrogen go when a plant decomposes. Hadas, et al., 2002 has experimental data from plants with C:N ratios from 11:1 to 136:1.
In summary, most carbon ends up in microbial biomass or as CO$_2$. If the process is allowed to proceed to infinity, then presumably, the microbial biomass will itself be decomposed by other microbes until all of it in the form of CO$_2$.
The nitrogen is mineralized, first into NH$_4^+$, then NO$_2^-$ and eventually all into NO$_3^-$ by various decomposing microbes. Normally these ions are attached to something else that they got from the soil and thus enter the soil. In the regular nitrogen cycle, the ions that eventually end up as NO$_3^-$ will be processed by other bacteria into atmospheric nitrogen (N$_2$). So, if decomposition proceeds to infinity, all the nitrogen would be expected to return to the atmosphere eventually too.
That leaves us with the small amounts of other biologically common elements such as phosoprous, calcium, etc. This will be the only stuff left. From Latshaw, 1934 (side note, not the oldest paper reference I've seen on Stack Exchange, but close!), and the chemical composition of a corn plant, considering all parts of the plant together, there is on average 44.9% oxygen, 44.0% carbon, 6.3% hydrogen, and 1.5% nitrogen by mass; all of these would be expected to leave as as gasses.
Of the remainder, representing 3.3% of the mass, probably most of that would stay. Whatever elements are left are most likely to form molecules with oxygen, since it is freely available in the air and likes to react with just about anything. Silicon, potassium and calcium don't seem to have any likely gaseous forms they could take, while sulfur (SO$_2$) and phosphorous (possibly P$_2$H$_4$) might take some more volatile forms. In the end, little will be left of the former plant, just a scattered residue of oxidized minerals, around 3% of the original mass.
