Feature Request: Add distribute_slack support to Linear Power Flow (lpf())

Description
Currently, distributed slack is supported for the non-linear power flow (n.pf()) via the distribute_slack and slack_weights parameters. However, these parameters are nonexistent for linear power flow (n.lpf()).

If a network has an active power mismatch prior to running n.lpf(), the entire imbalance is absorbed by the single slack bus. This can produce unrealistic power flows radiating from a single node. While the original feature request for distributed slack (Issue #10) mentioned both lpf and pf, PR 107 only implemented it for the non-linear module. Currently, the only workaround for linear power flow is to manually calculate the system mismatch and adjust the generator p_set values before calling n.lpf(), a feature which can be added directly to the lpf() function through adding a distribute_slack and slack_weights parameters to the linear power flow functions, mimicking similar behavior found in pf().

Proposed Solution
Bring feature parity to n.lpf() by adding the distribute_slack (boolean) and slack_weights parameters to the linear power flow API, mimicking the behavior already present in n.pf().

• Update the function signatures for n.lpf(), pypsa.pf.network_lpf(), and pypsa.pf.sub_network_lpf() to accept distribute_slack=False and slack_weights=None as default keyword arguments.
• If distribute_slack=True, compute and allocate the total active power mismatch to the active power injections of generators based on the provided slack_weights (defaulting to proportional to p_set or p_nom, matching the n.pf() logic).
• Proceed with the standard DC power flow calculation

All of this can be done manually outside of PyPSA, but adding it as a feature would streamline my workflow. Curious if others would find this feature valuable. I am happy to open a PR to implement the feature if so.

Thanks!

Start by reviewing the existing distributed-slack behavior in n.pf() and the implementations of pypsa.pf.network_lpf() and pypsa.pf.sub_network_lpf(). Trace how the linear power-flow entry points handle active-power mismatch, then add matching keyword arguments and verify that weighted allocation and the standard DC calculation behave consistently with n.pf().