While the use of dynamics with poor or slippery rail can create problems, used properly, dynamics can be a blessing when you need to control train speed carefully. The biggest advantage with dynamics over air is that you can make fine adjustments both increasing and decreasing dynamics whereas with air, once you have made a set, your only options are to either make a deeper set, or do a release. Once you do the release you must wait for the train to re-charge the cars before your airbrakes are again fully effective. Your airflow meter (if loco is equipped) is helpful here.
Too many sets & releases in too short a time can create a "no brakes" condition and the results of that can be spectacular!
New Engineers tend to set some air expecting to "feel" the results sooner than the brakes take effect. Then they set more air and, guess what, the train stops!
Sure, dynamics can fail, but it is rare. If your engine is running low on fuel the first sign is that both loading and dynamic braking will become erratic. Your ammeter will show this. If you have more than one unit in your consist, you can isolate the one that is having problems and continue to use the dynamics on the other units.
The P&L (today's W&I) with its roller coaster profile is a perfect railroad for dynamic brakes, and this goes back to NP's earliest use of diesels. The "Highball" was NP's manifest freight, making a round-trip from Yardley to Lewiston AND BACK in the 16 hours allowed in the good old days. In the steam era the "Highball" ran with two double-headed W-3 class Mikados. Of course this required two engine crews, making the "Highball" a prime candidate for diesels. This happened in 1951 with new 4-unit F-7's which had, guess what, DYNAMIC BRAKES. By 1957 dynamic brake equipped GP9's (NP 302-316, later BN 1740-1754) ruled the P&L.
Early pictures of the "Highball" with new F7's
In the mid 1960's the "Highball" often ran with a 4-unit FT set spliced with newer GP9s in the middle.