On their
"in depth" page, the Training Stress / Intensity graph and Training Power graph look exactly like their WKO counterparts. The "overview" looks like the performance manager. The predicted performance I'm not so sure about... it is basically just the green line from the overview,
they call "performance ability", which looks like level of fitness - fatigue, remarkably similar to TSB...
Both systems are based off of earlier work done by Bannister et al, hence the similarities.
They are claiming that predicts performance, which is a similar concept to TSB except in my experience a hugely positive TSB is about as bad as a hugely negative one.
This brings up one the biggest limitations to the performance manager, it doesn't attempt to predict performance whereas Bannisters impulse-response model did. Instead of predicting performance they came up with the TSB construct that models "adaptation to current training load."
In all fairness it is difficult to determine the various constants in Bannister's impulse response model for a given athlete and training program so the performance manager really is a good compromise. You get a fairly good impulse-response model but with much less effort. I have run both models side by side and they compare fairly well.
Raceday claims to determine the model constants on it's own but Dr Skiba hasn't been very clear on how exactly it does that. It also uses exponentially weighted moving average power instead of a 30s moving average to calculate its version of normalized power, which is more elegant and in theory could be better; however, in practice the difference is negligible.