false statement, false correlation
every active circuit in a amp will have a hard slew limit from device, bias and supply limitations - feedback or no and will distort for some value of "Transient" input
"soft" slew limiting, gain/phase changing with dynamic signal cause FM Intermodulation products and are also usually lumped in "TIM" too
"no feedback" amps have "TIM" distortion mechanisms, measureable "FM" Intermodulation Products
any nonlinear C, modulating gain working into a complex load will
and global feedback can be used without increasing "TIM" - just use a linear front end, attend to stage bias current demand, use 2-pole compensation
Bob Cordell clearly refuted Mati Otala's "high feedback amp" TIM analysis, prescriptions 30+ years ago - building a custom quadrature resolving IMD distortion analyzer and measuring "TIM" distortions down to 10s picoseconds in his high feedback, global loop gain
http://www.cordellaudio.com/poweramp/mosfet.shtml
more recently in AES conference papers Ron Quan has been revisiting TIM in audio op amps with different 3-tone test derived from RF amp practice and his own purpose built analyzer hardware - he had to look at generic 30 year old monolithic op amps, worse than the TL072 to get result above his instrument's noise floor
http://www.aes.org/e-lib/online/search.cfm