Minor Description
For some test files on the Adobe Acrobat Engineering website, Preflight output looks like this:
<preflight name="Disney-Flash.pdf">
<executionTimeMS>210</executionTimeMS>
<isValid type="">false</isValid>
<errors count="3">
<error count="1">
<code>1.0</code>
<details>Syntax error, Expected pattern 'obj but missed at character 'o'</details>
</error>
<error count="1">
<code>1.2.1</code>
<details>Body Syntax error, Expected pattern 'obj but missed at character 'o'</details>
</error>
<error count="1">
<code>1.2.1</code>
<details>Body Syntax error, Single space expected</details>
</error>
</errors>
</preflight>
Which suggests that Preflight doesn't correctly parse the objects. This is confirmed by a look at some of the offending PDFs in a hex editor, which reveals that the object identifiers in them are not terminated by any EOL character(s). AFAIK this is allowed in both PDF and PDF/A-1. More details + links to test files here ('Multimedia' table and below):
http://www.openplanetsfoundation.org/blogs/2013-07-25-identification-pdf-preservation-risks-sequel