There won't be any frequency rolloff. Ok, ok, maybe you'll suck it down from 100kHz to 90kHz with a long cable.
The 27k load will drop the level 1dB. On a scope it will look perfectly normal. However, sonically you might hear a slight loss of dynamics. A tiny bit of headroom compression. Nothing you can measure easily. Second order harmonics should start coming up. This is just the direction the sound will tend towards, which is easy to test by loading with 3k. At 50k there is no loss. At 3k you can hear it. At 27k I'm not sure, never checked. The change is gradual and I don't know exactly where it begins. Most likely you are safe. I'm sure many readers will be interested to hear your report back.
As far as overloading tubes, that would be a combination of signal swing and impedance. The stage operates at 5mA or 6mA (I forget). So a 2Vrms output (2.8V peak) into a 600 ohm load would be at the edge of clipping on negative going swing.
jh