[Task]: Make beam.Row() more robust for cross-language use

### What needs to happen?

TL;DR: There are two main issues here. The first is that there is no way to dynamically create beam.Row() elements to be used in a cross-language context. The second is that beam.Row() doesn't have type inferencing and defaults to `Any`, which is not understood by Java.

`StorageWriteToBigQuery` is a wrapper for a `SchemaAwareExternalTransform` that takes beam.Row()'s. The following creates beam.Row() elements from dicts:
```
output = (p
| beam.Create([{"num": 1}, {"num": 2}])
| beam.Map(lambda el: beam.Row(**el))
| beam.io.StorageWriteToBigQuery(table=table))
```

But this gives me `java.lang.IllegalArgumentException: Unknown Coder URN beam:coder:pickled_python:v1`. This is probably because of the `**` operator. Is there another way of creating beam.Row()'s dynamically?

I tried setting the args directly:
```
output = (p
| beam.Create([{"num": 1}, {"num": 2}])
| beam.Map(lambda el: beam.Row(num=el['num'])) <----
| beam.io.StorageWriteToBigQuery(table=table))
```

But looks like there is no type inferencing and instead defaults to `Any`:
```
java.lang.IllegalArgumentException: Failed to decode Schema due to an error decoding Field proto:

name: "num"
type {
nullable: true
logical_type {
urn: "beam:logical:pythonsdk_any:v1"
}
}
```

The following works fine (explicitly setting `int`):
```
output = (p
| beam.Create([{"num": 1}, {"num": 2}])
| beam.Map(lambda el: beam.Row(num=int(el['num']))) <----
| beam.io.StorageWriteToBigQuery(table=table))
```

### Issue Priority

Priority: 2 (default / most normal work should be filed as P2)

### Issue Components

- [X] Component: Python SDK
- [ ] Component: Java SDK
- [ ] Component: Go SDK
- [ ] Component: Typescript SDK
- [ ] Component: IO connector
- [ ] Component: Beam examples
- [ ] Component: Beam playground
- [ ] Component: Beam katas
- [ ] Component: Website
- [ ] Component: Spark Runner
- [ ] Component: Flink Runner
- [ ] Component: Samza Runner
- [ ] Component: Twister2 Runner
- [ ] Component: Hazelcast Jet Runner
- [ ] Component: Google Cloud Dataflow Runner