The idea is roughly right: a single type of string can look like particles with different quantum numbers like spin, charge, etc. (However, the proton is a composite particle while a single string should correspond to an elementary particle.) But there are a great number of technical issues to cover before it can be seen that quantum field theory, which is the framework of the Standard Model of particle physics, comes out as an effective description. That is what stringy physics should look like with the limited resolution and power of our "microscopes" (accelerators).
For example, the simplest type of string (embedded in spacetime) gives only particles with integer spin; to get fermions (spin-1/2, etc) the string must, loosely, be embedded in a "superspace" which has commuting and anticommuting coordinates. Also, there are fundamentally two types of string: open | and closed O, which have different particle spectra. And so on...
The underlying idea, though, is always the following:
When the *quantum mechanical* string theory is considered, there are states "|a>" and operators "A" just as in any quantum theory. The different operators A_1, A_2,... of the theory are building blocks, carrying a basic set of quantum numbers; each time you apply an operator to a state such as A_1|a(0)>=|a(01)>, the quantum numbers of the state change (and you say the string is in a different vibrational mode). Each of these states corresponds to what we'd see as a particle with a distinct spin, charge, mass etc. (the extended nature of the string being unobservable due to the limits of our "microscope" accelerators). In this way, acting with the full set of operators in all possible combinations on the "ground state" of the theory, you generate all possible quantum states |a(0)>, |a(01)>, |a(02)>,... of the theory, hopefully including all of the observed particles of the Standard Model.
However, the mass of these "excited states" is quite large and so none of the masses of the *observed* particles can come from these excited strings. That is, generally there are light particles predicted by string theory that we don't see in nature, as well as *infinitely many* very massive states that we wouldn't *expect* to see in the near future. It is the superfluous light states that are troublesome in constructing realistic models.