Fixup JointStiffnessController to output actuation instead of generalized_force

[RussTedrake]

Follows #22315, which implemented the same fix for InverseDynamicsController.

Previously there was a tacit assumption that the actuators were declared in the same order as the joints. This PR resolves that, and (to make this clear) renames the output port from generalized_force to actuation.

+@sammy-tri for feature review, please?
+@sherm1 for platform review, please?

---

This change is 

[sherm1]

Out of order platform (still need feature review)
Q: in #22315 you retained a generalized_force port in addition to the actuation port. Here you are replacing the generalized_force port with actuation. Should these be treated identically? Also with this change I believe the release notes should say "breaking change" since the deprecated port has changed behavior.

Reviewed 7 of 7 files at r1, all commit messages.
Reviewable status: 1 unresolved discussion, LGTM missing from assignee sammy-tri(platform)

---

systems/controllers/joint_stiffness_controller.h line 27 at r1 (raw file):

 * and `τ_app` contains applied forces from force elements added to the
 * multibody model (this can include damping, springs, etc. See
 * MultibodyPlant::CalcForceElementsContribution()). B⁻¹ is the inverse of the

nit: inverse (or pseudoinverse if underactuated)

[RussTedrake]

The argument for also exposing generalized_force in the inversedynamicscontroller was because it can be useful in different workflows (like actuating a floating hand, which is a pain to add actuators for). i couldn't imagine any similar motivation here... it's joint stiffness control, afterall. So I intentionally made a different decision here.

Reviewable status: 1 unresolved discussion, LGTM missing from assignee sammy-tri(platform)

---

systems/controllers/joint_stiffness_controller.h line 27 at r1 (raw file):

Previously, sherm1 (Michael Sherman) wrote…

nit: inverse (or pseudoinverse if underactuated)

the same block of texts says this class assumes the plant is fully-actuated.

[sherm1]

Reviewable status: LGTM missing from assignee sammy-tri(platform)

---

systems/controllers/joint_stiffness_controller.h line 27 at r1 (raw file):

Previously, RussTedrake (Russ Tedrake) wrote…

the same block of texts says this class assumes the plant is fully-actuated.

Oh, I didn't read the part that hadn't changed -- sorry! All good.

[jwnimmer-tri]

I agree with the reasoning for deprecating the generalized_force port, in contrast with #22315 which keeps it around.

However, @sherm1 with your platform hat on, please help us defend our promise to uses:

On very rare occasions it is impractical, too expensive, or logically unsound to meet the full three month deprecation window. In these cases the change will be announced in the “Breaking changes” section of the release notes.

-- https://drake.mit.edu/stable.html

Breaking changes should be "very rare". Any PR which is a breaking change needs stiff scrutiny from the platform reviewer. For any breaking change, the platform reviewer should ask the author to explain their rationale for not doing a more gentle deprecation. If there is a good reason, that's fine -- we just want it written down as part of the review.

In this case, it seems to me like we can copy-paste the old calc method for the deprecated output port to serve as the deprecated calc's implementation, and so it seems to me like introducing a non-breaking deprecation would be practical and sound.

[sherm1]

I see your point, Jeremy. WDYT Russ?

Reviewable status: LGTM missing from assignee sammy-tri(platform)

[RussTedrake]

Done. I agree that's better (and had considered doing it in the first place).

Reviewable status: LGTM missing from assignee sammy-tri(platform)

[sherm1]

Reviewed 3 of 3 files at r2, all commit messages.
Reviewable status: LGTM missing from assignee sammy-tri(platform)

[sammy-tri]

Reviewed 4 of 7 files at r1, 3 of 3 files at r2, all commit messages.
Reviewable status: 1 unresolved discussion

---

systems/controllers/joint_stiffness_controller.h line 159 at r2 (raw file):

  // Calculator for the deprecated output port.
  void CalcOutputForce(const Context<T>& context,
                           BasicVector<T>* force) const;

nit: odd indentation