Improve naming of some QEE HLT2 lines

FYI @jiuzhao @hyin @kaaricha

Just some renaming of some HLT2 QEE lines.

• Add "HighMass" to name of same sign Z->ee (mu mu) lines to differentiate them from the lower-mass dilepton lines we have in QEE.
• Removed "prescaledFullEvent" from the names of the full-stream DiMuonNoIP lines. The "FullEvent" is no longer required, as all our Full-stream lines now have "Full" in the name. I also removed "prescaled" - I didn't think this was necessary in the name but shout if you disagree.

Also added a small warning in spruce_qee.py in the hlt2_filter_codes: I realised after I finished off !2743 (merged) that the regex won't work for Hlt2QEE_DiMuonNoIP_prescaledFullEventFull or Hlt2QEE_DiMuonNoIP_prescaledFullEvent_ssFull on samples before !2743 (merged), as there are two "Full"s in the name, both of which will be replaced, whereas we'd want to only replace the final one. Fortunately there is no sprucing line for these (maybe that should be followed-up - FYI @kaaricha), so no issue, but if one is added, then the filter code won't work through _hlt2_decision_regex on 2023 or earlier samples. I think you'd just silently get 0 rate through the line in that case. A regex that doesn't go through _hlt2_decision_regex would be needed in this case. Avoiding this in the future is part of my motivation for changing the name of these HLT2 lines.

Bandwidth changes

Using the instructions provided in the CodiMD with 10k events

• Since the DeterministicPrescaler sets it seed based on the line name, the number of events triggered by the two prescaled DiMuonNoIP lines change in a small-scale test, which causes a spurious change in bandwidth at the ~+30MB/s level to Full-stream on tape in this test. There's no change at sprucing, which is unsurprising as these HLT2 lines don't have sprucing lines keeping their output -> disk.
• Bandwidth and rates of other lines are verified to be unchanged. See more in discussion.