Update T1 and T2* with the new curve analysis class

[yaelbh]

Now that #46 has been merged, we can simplify T1 and T2*

[yaelbh]

@nkanazawa1989 Doing this for T1 is not straightforward. For example:

• The function _initial_guess doesn't have an access to user_p0. This is required for parameter c, whose initial guess is a function of a (either user_p0["a"] or the default initial guess). This can be overcome by ignoring c inside _initial_guess and dealing with it only in _setup_fitting, but a nicer solution is preferable.
• We need to access the metadata to get the conversion factor. I think this is handled by one of the open PRs.

I'll postpone this task to after the release.

[nkanazawa1989]

The _initial_guess function is a helper for RB experiment and it's necessary defined for others (this is defined because IRB also use the same code). I think the implementation of spectroscopy analysis would help the migration.

Regarding the circuit metadata, it is now available by #99. Alternative approach would be adding "unit" to analysis option and use self._get_option("unit") in analysis. I prefer latter one, because unit doesn't change per circuit.