I'm not good on the png thing vs pdf's. However, you may have missed Bob's posting on some of the history of why USATF uses png as opposed to any other format(see at end of this responce).
Maps that I scan depend on the Certifiers making sure they are done well. I have told several that the maps sent in are poor and must be redone. I feel as a whole things have improved.
The West VC has gone all electronic and I'm working on the East VC going in that direction. I have some Regional Certifiers in the East doing electronic, but it's a work in progress.
Here are Bob's comments about why PNG's are the best format!
From: Bob Baumel
Re: Format used for our maps on the USATF site and how it came about!
First, the question of raster or vector formats. To the extent that we need to scan hard-copy maps, raster is the only option. Vector formats are possible only for computer-generated maps, but only if the map is generated with the specific intention of making it fully scalable (I'm sure that, in practice, a lot of computer-generated maps aren't fully scalable, but contain combinations of vector and raster elements). And for a map to remain fully scalable, it must remain in electronic form throughout the process. If it ever gets printed out, requiring us to scan the hard copy, the result will be a raster image, and we'll have no way to get back to the vector graphic.
There's also an issue of file formats for vector graphics. Most vector graphics are produced in proprietary formats, such as Adobe Illustrator. We certainly won't adopt a proprietary format of that sort as our standard. There is a non-proprietary vector format that's been under development for many years called SVG (Scalable Vector Graphics), but I doubt that it's supported widely enough to consider using it as our map standard.
Anyway, given that we needed to choose a single format as our standard, and given that we definitely need to support scanned hard copy maps, it had to be a raster format. Given that we needed to pick a raster format, the optimal choice was clearly PNG, which is superior to older raster formats such as GIF, JPEG and TIFF. The PNG format is lossless and non-proprietary -- and among all lossless raster formats, it provides the best compression.
Note that aside from choosing PNG as our format, we also made choices involving color and resolution; it needs to be monochrome (black & white) and 300 dpi resolution, so an 8.5" x 11" map is 2550 x 3300 pixels.
JPEG is a "lossy" format, which loses quality every time the image is edited and re-saved. JPEG tends to be best for photographic images, not maps. Our choice was definitely the lossless PNG format.
What about PDF format? Every time somebody suggests PDF, I must emphasize: PDF IS NOT A GRAPHICS FORMAT. It's a file format that can contain all sorts of stuff, including text, vector graphics and raster graphics. Simply because a map was saved in a PDF file doesn't mean that it's scalable. The graphics inside that PDF may be raster images, of any resolution (so when you "zoom in" within that PDF, they may look terrible, full of "jaggies"). Thus, if we simply say that we want maps in PDF format, we haven't said anything about image resolution.
People often produce PDF files by initially generating a document in some proprietary format (such as Microsoft Word for text) and then "printing" to a PDF file. In the same way, if they start with fully scalable graphics generated in a proprietary program such as Adobe Illustrator, and then convert to PDF, they will indeed get a PDF with fully scalable graphics. But even in this case, if a measurer sends such a PDF to a Certifier, I suspect that many of our Certifiers will have a difficult time adding the necessary annotations (Remember that the Certifier needs to write in the assigned course number and dates of certification validity) and then preserving it as a fully scalable PDF.
Another issue involving PDF files is that our maps are posted on the USATF site where they must be easily available for both screen display and printing. If we post PDF files, screen display can be problematic. Web browsers cannot display PDF files directly, but require a plug-in (usually Adobe's PDF plug-in) to display them in a browser window. Displaying a PDF file this way (using PDF plug-in) tends to be slower than direct display of a PNG graphic by the web browser. More seriously, the required plug-in isn't even available for all web browsers on all platforms. For example, on the Macintosh platform, people running the popular Firefox browser cannot view PDFs in a browser window (mainly because Adobe hasn't seen fit to provide its PDF plug-in for any Mac browser except Safari); thus, Mac users running Firefox would be forced to download the PDF file for every map they wish to view.
PNG files are handled directly by all modern browsers and work well for both screen display and printing. Along these lines, it's interesting to recall some history. Pete referred below to discussions between Stu Riegel, Keith Lively, myself (and Pete), in which we chose the current PNG format. Those discussions took place in early 2006. However, as Mike indicated, I had extensive correspondence with Keith several years earlier regarding the methodology for posting and displaying maps on the USATF site. Interestingly, even at that early date, we decided to post only PNG files on the website (which might have been considered somewhat "bleeding edge" at the time, since some of the browsers still in use didn't support PNG very well). The techniques adopted then weren't as efficient as possible: Keith posted two PNG files for each map -- a low-resolution version for screen display and somewhat higher resolution version for printing. Also, RRTC wasn't sending the data to Keith in PNG format. As Mike described, he scanned maps into very high resolution TIFF files, which (I believe) weren't compressed at all, so were huge files. Mike sent those big TIFF files to Keith, who ran a script to convert them into the two PNG files that Keith posted on the website.
When several of us revisited the issue in early 2006, we simplified the procedure greatly, by selecting a single PNG format that could be scanned directly by the RRTC Registrar, posted on the USATF site without any conversion, and used for both screen display and printing. At 300 dpi, this new PNG standard was somewhat higher resolution than both of the files Keith was posting previously, but it was a single file that replaced two files, and by keeping it monochrome, file sizes were small enough to work well on the web.
The format selected in 2006 is probably still optimal, given that we can't avoid working with scanned hard-copy maps, which require a raster format -- and PNG remains the best raster format. People periodically say we should use some other format, and the format mentioned most often is PDF. But as I've said, PDF isn't a graphics format. PDF files can contain all sorts of things. PDF files are extremely useful for transmitting information, precisely because they can contain so many things. But we can't base our map standard on the PDF format. Saying that a file should be in PDF format says nothing about the nature or resolution of the graphics inside the file. It may contain scalable graphics, but it might contain nothing but low-resolution raster graphics. And for screen display of maps on the website, we must remember that not everybody can view PDF files in a browser window.