Allow data-free compression with 3D weights (#3755)

### Changes

Update the condition introduced in
#3706 to allow data-free
compression with 3D weights.

### Reason for changes

Only data-aware compression is not possible. Caught with optimum-intel
tests
https://github.com/huggingface/optimum-intel/actions/runs/19623687915/job/56188570802.

Author: nikita-savelyevv

src/nncf/quantization/algorithms/weight_compression/algorithm.py

@@ -869,12 +869,17 @@ def get_weight_compression_parameters(
                         and ov_version
                         and version.parse(ov_version) <= version.parse("2026")
                         and node.metatype in self._backend_entity.matmul_metatypes
+                        and (
+                            self._data_aware_compression
+                            or self._awq
+                            or self._sensitivity_metric == SensitivityMetric.HESSIAN_INPUT_ACTIVATION
+                        )
                     ):
                         # MoE operations are usually matmuls, so the check for matmul metatype is done
                         # This is to avoid raising the error for non-MoE cases with 3D weights.
-                        msg = f"""NNCF does not support 3D weights with current version of Openvino {ov_version} 
-                                due to a known issue in statistics collection Ticket - 176465
-                                Node with weight: {node.node_name}"""
+                        msg = f"""NNCF compression algorithms do not support 3D weights with current version of 
+                                Openvino {ov_version} due to a known issue in statistics collection Ticket - 176465. 
+                                Node with weight: {node.node_name}."""
                         raise nncf.UnsupportedModelError(msg)
 
                     if self._backup_mode != BackupMode.NONE: