Version 2 of PEARLTransfit with improvements to in/output properties

[RichardWaiteSTFC]

Is your feature request related to a problem? Please describe.
Following on from work in #39101

Have identified (with Nick Funnell on PEARL) some breaking changes we would like to make behaviour of PEARLTransfit more intuitive

• Have T\S_Fit workspaces as output properties - this would stop workspaces being overwritten and give workspace history
• Have S_Fit workspace as an input property required if Calibration=False rather than hard-coding name
  • This would also enable sequential fitting by passing the results form previous temperature
• Have sample temperature and other debug info in an output table rather than printing the info to the log
• Get rid of Ediv parameter - it's not ideal to bin with constant bin width and this has unite eV where data is in meV!
• Get rid of Bg0Guess etc. - allow user to read in parameters from an input table like the do when Calibration=False

A nice to have would also be to support sequential fitting - so passing multiple files, however sometimes files need to be summed as correspond to the same temperature. Would need to distinguish when runs need to be summed or fitted separately (e.g. + or , delimiters could be used as in powder diffraction scripts? The behaviour above means a sequential fit could be implemented by passing the previous T_Fit table as the calibration for the subsequent temperature.