Hi Dave,
As one who has found something similar on a past bicycle and then had to deal with it, I can't help but offer a bit of sympathy and empathy. A person just feels kinda sick, looking at it.
Now, to causes and solutions...
Yes, I agree with Robin that recent environmental legislation worldwide has made paint/finish preparation not what it was, and it is important to touch-up paint and convert rust soonest. Local conditions can play havoc as well -- high humidity and salted roads are just deadly to a steel frame, and can get the tinworms started tunneling under the paint. Fenders generally help a lot (provided they don't rattle against paint at their edges), as does making sure all cable housings are tied-down and riding against scuff pads.
However, where your photos all show rust around the lugs and and brazed bits (see the chainstay bridge at the fender-mounting boss), it makes me wonder if the original problem might have been due to inadequate rinsing/washing of the flux used in brazing. That was the cause of lug-proximate rusting on an other-brand frame I owned, and there were no other exacerbating or contributing factors except generally poor paint (20 years after purchase, the clear-coat never set-up and would still become gummy after the bike sat in the sun on a hot day). Paint tends to be thinnest on the edges of the pointy bits, leaving those same edges especially vulnerable. It is an area of frame preparation I watch ever so carefully as a hobbyist framebuilder. I've tried a gasfluxer when brazing, which pretty much solves the problem before it starts, but anytime a powder-mixed paste flux is used -- with the higher temps needed for brass, especially -- one has to really watch it, as it can harden to a clear, glass-like surface that can later rehydrate beneath the paint and cause problems. It can be hard to see, and it really takes boiling water to remove it properly in a short timeframe.
I agree, the "cure" is to remove as much rust as possible (a soft-media blasting using something like broken walnut shells would prevent surface-hardening of the steel while removing any loose rust). Later treatment with a rust converter (my favorite is Ospho
http://www.ospho.com/ ) will not only stop any remaining rust, but prepare the surface with a light etching so paint can better bond to the surface.
If your frame was powder-coated, that has its own perils as well. Now often a necessity because of those same enviro-regs, melted plastic sometimes just does not bond as well as wet-paint. For example, the first time I removed the rear wheel on my Sherpa, the paint lifted off the
entire inside-left dropout face in one sheet, fluttering to the ground with the wheel. A quick swipe of Ospho on the unprimered bare steel, an overnight dry, then a retouch with semi-flat Krylon over primer, and I'm good to go. I went with the satin black option so I could quickly make near-invisible finish repairs on-tour, if necessary (I tuck in my little bottle of touch-up paint as part of my regular touring kit).
The good news (yes, there is some!) is the steel doesn't look at all bad in your photos -- some surface rust yes, but no bubbling or flaking I can see. If you can get on this sharpish, I really think the frame is not imperiled and a respray in part or whole will make it fine. If a professional resto and respray is out of the question, it is possible to do a home-based prep and wet-paint respray and still have a durable paint job, even with less-aggressive paint formulations. I took on that task with my tandem, which had as much bare metal as paint when I bought it used. It's been close to 20 years now, and my respray shows no chips or rust. Same with the others I've painted.
Hope this helps, Dave. A real shame this happened, but really good chances for a complete recovery of the patient with a long and healthy lifespan ahead.
Best,
Dr. (not the medical kind!) Dan.
