Failure to collect operations that join two lazily retrieved Polars lazyframes

### What happens?

I have downstream functions that expect multiple `polars.LazyFrame` objects. In my upstream code, I want to source my data from DuckDB to enable efficient pushdown. However, using `DuckDBPyRelation.pl(lazy=True)` raises a `ComputeError` when operating on multiple `LazyFrame` objects.

### To Reproduce

```python
import duckdb

with duckdb.connect() as con:
con.sql("CREATE TABLE tbl (t1 INTEGER)")
con.sql("CREATE TABLE tbl2 (t1 INTEGER)")

# Retrieve the results
rel = con.table("tbl").pl(lazy=True)
rel2 = con.table("tbl2").pl(lazy=True)

# Compute and raise the error
df = rel.join(rel2, on="t1", how="inner").collect()
```

the block above raises

```
ComputeError: caught exception during execution of a Python source, exception: InvalidInputException: Invalid Input Error: Attempting to execute an unsuccessful or closed pending query result
```

### OS:

Linux - aarch64

### DuckDB Package Version:

1.5.5

### Python Version:

3.12

### Full Name:

Akshay Gupta

### Affiliation:

Capital One

### What is the latest build you tested with? If possible, we recommend testing with the latest nightly build.

I have tested with a stable release

### Did you include all relevant data sets for reproducing the issue?

Yes

### Did you include all code required to reproduce the issue?

- [x] Yes, I have

### Did you include all relevant configuration to reproduce the issue?

- [x] Yes, I have

Start at DuckDBPyRelation.pl(lazy=True) and reproduce the failure with the two relations joined through Polars, then follow the execution path into rel.join(...).collect(). Done means the supplied example collects successfully and returns the joined result without a ComputeError.

Written by the indexing model from the issue text.