Telemetry: fix reporting of profiling and appsec state

[p-datadog]

What does this PR do?
Fixes telemetry reporting of profiling and appsec enabled/disabled state (in app-started events).

Profiler was reported as not running in the app-started event regardless of actual state, due to the way enablement check was implemented - it looked at whether profiler was started, not just enabled, and profiler was started after telemetry by the component tree.

This PR changes the check to consider whether the Profiler instance exists, which should indicate whether profiling is actually enabled and working.

AppSec was reported based on its configuration, not actual state. If AppSec was requested to be enabled but failed prerequisites (Ruby version/FFI version) or failed to initialize, it would still be reported as enabled.

We have the configuration for each component sent separately which contains the requested enabled/disabled values, therefore the product field should have the actual state.

I took the liberty to change AppSec.enabled? to read component state also.

Other, smaller changes while I was looking at the code:

Telemetry is now shut down last, to permit all other code to send diagnostics during shutdown. Previously process discovery was shut down last, it didn't send anything via telemetry that I saw, but I moved it ahead of telemetry.

Motivation:
Code review in #5067

Change log entry
None

Additional Notes:

AppSec does not store the settings into its component, therefore all configuration is not necessarily correct (during reconfiguration, existing components will report/use new configuration rather than "their" configuration). This most notably makes rasp_enabled? code weird in this PR, but applies to all of AppSec configuration from a quick look.

AppSec initialization block did not report the exception it rescued, I changed it to do so (the tests I wrote didn't report any reason why AppSec failed to initialize without this change).

Only profiling reports the reason why it wasn't enabled, DI and AppSec do not.

How to test the change?
Integration tests added to verify both configured enabled/disabled state and actual enabled/disabled state for all products reported to telemetry (AppSec, DI, profiling).

[github-actions]

Thank you for updating Change log entry section 👏

^{Visited at: 2025-11-20 17:51:39 UTC}

[github-actions]

Typing analysis

This PR does not change typing compared to the base branch.

[pr-commenter]

Benchmarks

Benchmark execution time: 2025-12-02 15:43:20

Comparing candidate commit 1fc3d9f in PR branch telemetry-profiling-reporting with baseline commit 895d80e in branch master.

Found 0 performance improvements and 1 performance regressions! Performance is the same for 43 metrics, 2 unstable metrics.

scenario:profiling - Allocations ()

• 🟥 throughput [-502318.866op/s; -495616.334op/s] or [-15.127%; -14.925%]

[Strech]

I still have gaps in understanding of the life-cycle of our components and especially telemetry, but looks good to me (I would recommend waiting for someone else feedback too).

Maybe few improvements over typespecs

[p-datadog]

So this PR is on top of #5074 and most of the diff actually was (and was reviewed) in that PR.