InfluxDB Cloud csv.form() does not accept dateTime:number format

Nobody has claimed this yet.

__Steps to reproduce:__
List the minimal actions needed to reproduce the behavior.

1. create a .csv file with this content:
```
#group,false,false,false,false,false,false
#datatype,string,long,string,string,long,dateTime:number
#default,to1,0,analogvalue,testsubject1,,
,result,table,_measurement,_field,_value,_time
,,,,,0,1482669077000000000
,,,,,-1,1482669078000000000
,,,,,-1,1482669079000000000
,,,,,-1,1482669080000000000
```

2. Upload that in the cloud
3. Log into your InfluxDB Cloud account
4. Choose "Explore" in the menu on the left
5. Type this in the script editor:

```
import "experimental/csv"

csv.from(url: "https://your-s3-bucket.s3.eu-central-1.amazonaws.com/your-csv.csv")
|> to(bucket: "test-bucket")
```

__Expected behavior:__
I expect the data inside the .csv file to be saved into the InfluxDB bucket called test-bucket.

__Actual behavior:__
After clicking the "Submit" button i get:
error in csv.from(): parsing time "1482669077000000000" as "number": cannot parse "1482669077000000000" as "number"

However if i use dateTime:RFC3339 datatype and time data like 2018-05-08T20:50:00Z then the data gets imported into the InfluxDB bucket.

I am recording analog signal with a high sampling rate and add timestamp data later on in google sheets. Problem is google sheets can not auto-increment RFC3339 formatted timestamps at sub-second levels (have several thousand rows, so manual entry is not an option 😉 ). Google sheets would easily autoincrement nanosecond values...but for some reason InfluxDB Cloud does not consume dateTime:number formatted data. 😕

__Environment info:__

InfluxDB Cloud deployment in Frankfurt on AWS.
CSV file written with Notepad on a Win10 x64 machine
Using Chrome V89.0.4389.90

This bug combined with #21011 prevents me using InfluxDB Cloud for storing my time-series.

Hints (if any) are highly appreciated!

Start at the experimental/csv csv.from entry point and reproduce the import with the supplied CSV using dateTime:number, then compare it with dateTime:RFC3339. Done means numeric nanosecond timestamps are accepted and written to the target bucket without the parsing error.

Written by the indexing model from the issue text.