The Fermi level is your reference binding energy, so if you have entered the correct sample bias voltage into your software, the Fermi level is at EB=0eV.
If your software doesn't calculate binding energies automatically, you have Ekin,F=hv+eUBias
1) You clearly have not included your sample bias in your evaluation. Please shift the whole spectrum towards higher energies by the value of your bias. That's what I meant above by Ekin,F=hv+eUBias, right now this is just Ekin,F=hv, so the spectrum is simply false at the moment.
2) After that, you do not need to put any fitting lines through your spectrum. The 0eV IS your Fermi energy assuming that you don't have charging effects, that's what I meant by my very first sentence above.
3) A cutoff analysis as the one you are performing there is used for determining the ionization energy IE of your sample, but of course you have to use the lowest-bound component, not the second highest. It is IE=Ecutoff+WF, in which WF is the work function.
4) The work function is determined from the middle position of the secondary electron edge, i.e. the cutoff at high binding energies. It is WF=hv-ESEC. Also, the WF can be used to check whether you have charging in your system, so the deviation from literature values shouldn't be too big.