Thank you for the compliment!
I do my finish rendering and such in either Phocus or Lightroom. If Phocus, I bring the result into LR then. I use the Slideshow module to create the drop-shadow framing I want. I create JPEG files to export from that (option key changes the 'export to PDF' button to 'export to JPEG') and set the output to be 3000x3000 (for 33x33 square crop) or 3000x2250 (for 44x33 full frame) pixels dimension. These files are imported back into LR and the IPTC metadata is copied from the master file into the JPEG for each, then metadata is saved to the files.
At this point, the files are uploaded to Flickr.com. For each photo, I copy the BBCode output from the Share button with the output parameter set to 1600x1600 or 1600x1200 (about) for posting here. I edit out the unnecessary information from the default BBCode output when I create the post to display the file here (or anywhere else on the web).
Using this process means that my files are all located in ONE place online no matter which forum or group I'm sending them to, and i can control the size of the display image with very good granularity as well as give access to half-resolution images through flickr directly for those who would like to see larger sized versions.
The process is dependent upon my understanding of just how much input and output sharpening I need to apply to a 6000x6000 or 6000x8000 image file in order that a 3000x3000 or 3000x2250 file looks good, of course, and I can only point to lots and lots of time (20+ years) and experience sharpening my photos for that.
G