Like kvflyer, I have restored many of the Scott amplifiers and tuners (own a few of those as well). I have replaced many of the Ceracaps, but have never found a bad one (as tested on my Sprague TO-5). They are Mylar (polyester) caps, not paper. If I find any of the molded plastic paper caps (eg. Pyramid), I always replace those. They always test leaky. I have read about a case or two where a Ceracap lead came loose, leaving the cap open, but this is evidently rare. I have not seen that failure mechanism.
The 25 uF 25V sections on one of the cans (cathode bypass cap sections) always test bad and I don't trust the old cans at all. On an early 299B restoration, it seemed to be working fine and the cans checked OK for ESR, but one day I heard cracking and popping and one of the cans was very hot and on the verge of exploding. Since then, I don't trust any of the 50-60-year old cans and replace all of them, either with properly valued CE new caps, F&T or JJ HV caps, or Hayseed Hamfest. For the can in the bias/12V heater circuit, I usually disconnect these and use 100uF/100V axial or radial modern caps installed underneath, leaving the disconnected can in place.
Another must replacement is the selenium rectifier bridge which powers the output tube bias circuit and the four phono and line stage 12AX7 heaters. These go bad gradually and began produce lower and lower voltage and should always be replaced with a new silicon bridge rectifier. These produce a bit higher voltage than the original selenium bridge, and you need to increase the value of the first dropping resistor after the bridge to around 30 ohms (3-5W).
Check also the 18 ohm 1W resistors in the heater/bias circuit string. These often rise in value and need to be replaced.
To me these are the minimum things requiring attention when acquiring a vintage Scott amplifier and it's worth the expense and work--they are excellent amplifiers.
Dave