download with lynx

Edited text of question

  Always?  I don't think so.  This is certainly not typical of browser behaviour (binary transfers are not its main intended purpose).  Having a browser default to making transfers that it won't then display sounds rather odd when you think of its intended purpose, and the designers have not included this as an option.  They programmed the appropriate functionality into Lynx, assigning responsability to web page owners to properly name their files. (proper extentions)
  If you want to make binary download SIMPLER, then writing a simple script which could accept an URL name and then create a download HTML page with a link which is "d"ownloadable would be easy, and somewhat more convenient.
  Lynx does not include an option to download binary and then display the results, so if you were able to set it up to do binary transfers as default, it would not display HTML pages - just download them.  This kind of defeats the purpose of a "browser", even though it lets you download errantly named files in binary mode.
  Errantly named files are not too common, so it shouldn't be all that much of an inconvenience when you encounter them.  The best thing to do, I would think, is to try to make the process a little less painful via scripts, etc...

Edited text of question

I know you do not want HTML pages to download that way, that's just how it might have to be if binary transfers were the default.  "D"ownloading should always be binary, but it isn't in your setup?  I haven't encountered this before... even if I download a TEXT file it does it binarily with the "d"ownload function.

Edited text of question