Fix timeout of duplicated sessions

[sjanusz-r7]

The PR here: #16369 allowed us to run sessions -u {session_id} on a Meterpreter session to 'duplicate' a Meterpreter session (read that as open a new Meterpreter session with a default payload that is decided by the shell_to_meterpreter.rb script).
On Python, this has resulted in the initial session waiting for the newly spawned session to exit, thus MSF Console thought the session has been timing out.

This PR:

• Fixes the above issue by running the last call to cmd_exec in transmit_payload as non-channelized, meaning the initial session does not wait for the process to finish.
• Adds in more possibilities to detect the Python binary on the remote machine.
• Defaults more cmd_exec calls to non-channelized.

Verification

• Start msfconsole -q
• use payload/python/meterpreter/reverse_tcp
• set your options and get a session
• sessions -u -1 to 'upgrade'/duplicate/spawn a new session
• Verify the initial session can still be interacted with
• Verify that killing the initial session does not kill the new 'duplicated' session

[adfoster-r7]

@msjenkins-r7 retest this please

[adfoster-r7]

Release Notes

Fixes a bug when running multi/manage/shell_to_meterpreter to upgrade from a Python Meterpreter session to a Native Meterpreter session would kill the original Meterpreter session