PDF is not a container format (unless you just mean that it can contain arbitrary contextless binary data, which it can, not that any PDF viewer will know what to do with it), and it can contain only a few different image formats (sort of; it's really just image compression algorithms). Notably, its only TIFF support is CCITT fax, which only really specifies 1-bit color depth. It also has a few JPEG encodings. Beyond that, it's straight raster images, potentially compressed with RLE, flate, or LZW. (To be fair, a TIFF is likely to also just be LZW-compressed.)

Specifically, most PDF creators give you very little (obvious) control over how an image is going to be imported. It might decide to use one of the lossy JPEG algorithms. Even if it imports it losslessly, it might decide to place it as an image inside a page, which would screw up formatting.

So, yes, a PDF could be used to keep an identical copy of the original TIFF data, but making that happen is problematic. If the intention is to edit the images, you don't want to use a program that might lossily compress the data, then edit it in another program that might lossily compress it again.

As a final rendering, yeah, it's fine, as long as the results look okay to you. Using it as an intermediary image, though, is fraught with downfalls.