Several improvement suggestions for the MQTT Bridge documentation

File: [/en_US/data-integration/data-bridge-mqtt.md](https://www.emqx.io/docs/en/v5.1/data-integration/data-bridge-mqtt.html)

When I need to establish a bidirectional MQTT bridge, EMQX currently supports three configuration methods:
1. Create a bridge on a node and configure ingress and egress at the same time
2. Create bridges on both nodes and configure egress on both nodes.
3. Create bridges on the two nodes and configure ingress on both nodes.

But now the documentation fails to tell me what is the difference between these configuration methods. What scenarios are they suitable for?

For example, when I need all clients to publish messages to the topic `bridge/topic` and subscribe to `remote/bridge/topic`, which configuration method should be used? Or what are the requirements for topic design of different configuration methods?

Start with /en_US/data-integration/data-bridge-mqtt.md and review the three bidirectional MQTT bridge configuration methods described in the issue. Document how they differ, which scenarios suit each method, topic-design requirements, and which method fits publishing to bridge/topic and subscribing to remote/bridge/topic. Confirm the examples are clear and actionable for readers.

Written by the indexing model from the issue text.