[QEE] Fix mistakes with Hlt2QEE_ZToMuMu_SingleNoMuID (prescale + decay_descriptor)

In !2309 (merged), a prescale of 0.1 was added to Hlt2QEE_ZToMuMu_SingleNoMuID, and it should be removed. This line is intended for efficiency studies in the electroweak analyses.

On 25k events of upgrade_minbias_hlt1_filtered (Rate: 1.650 MHz), this line didn't fire. This gives a bound of <66Hz for the rate.

The efficiency on Z->mumu MC simulation, with HltEfficiencyChecker's CanRecoChildren denominator is 0.959 +/- 0.008

---

@rjhunter
Found an incorrect decay descriptor in the SingleNoMuID line, fixed now. Z0 -> mu+ mu- to [Z0 -> mu+ mu-]cc.
This is necessary as the SingleNoMuID has an 'asymmetric' set of muons, i.e. one ismuon_long_muon and one long_muon. With the incorrect decay descriptor, the ordering is fixed and the Probe muon would always be forced to be mu-. Now with this changes it is fixed.

Efficiency is expected to change. (Seems MooreAnalysis is broken by the nightlies, so until that's fixed):
Using the Moore Job Log (on Zmumu MC) without any reconstruction denominators like CanRecoChildren we see a small increase in efficiencies.
With cc: LAZY_AND: Hlt2QEE_ZToMuMu_SingleNoMuID #=1074 Sum=328 Eff=|( 30.54004 +- 1.40540 )%|
Without: LAZY_AND: Hlt2QEE_ZToMuMu_SingleNoMuID #=1074 Sum=304 Eff=|( 28.30540 +- 1.37460 )%|

When these lines were added, the rates were too small to be seen. With these changes: over all the events (15k) in upgrade_minbias_hlt1_filtered (1.65 MHz), no events fired. Thus Nominal rate estimate < 110 Hz.
Over 25k events in hlt1BWdiv (0.95 MHz), no events fired. Thus '23 Rate estimate < 38 Hz`.

The rate still is too small to show up on Minbias, seems fine. Some small increase in efficiency is expected as the mu- Tag, mu+ probe option is now possible.