I had a similar experience with JJ 7951s, although reduction in output power wasn't as severe. Electroharmonix 7951s were much closer to original data sheets.
The JJ data sheet charts are at lease a honest account of the tubes lack of ability compared to NOS 7591. Have to give them credit for that.
The JJ charts do show that what they are selling is not really a 7591 but something lessor in ability.
For example...
The Sylvania data sheet shows with Vg2=300V, Vg1=0V and Va=50v results in Ia=175mA, the JJ only Ia=90mA.
The Sylvania data sheet shows with Vg2=300V, Vg1=0V and Va=70v results in Ia=195mA, the JJ only Ia=130mA.
The JJ 7591 has peak anode current levels are not even at NOS 6BQ5 levels (Ia=140mA with Vg2=250, Va=75V, Vg1=0), not really 7591 performance levels at all.
Unfortunately I expect there are a lot of users that do not read and compare the charts carefully but simply go with matching the part number.
I know I did at first and only later really understand what JJ is selling is more of a 7591 "inspired" tube that may work in some applications.
The JJ charts do show that what they are selling is not really a 7591 but something lessor in ability.
For example...
The Sylvania data sheet shows with Vg2=300V, Vg1=0V and Va=50v results in Ia=175mA, the JJ only Ia=90mA.
The Sylvania data sheet shows with Vg2=300V, Vg1=0V and Va=70v results in Ia=195mA, the JJ only Ia=130mA.
The JJ 7591 has peak anode current levels are not even at NOS 6BQ5 levels (Ia=140mA with Vg2=250, Va=75V, Vg1=0), not really 7591 performance levels at all.
Unfortunately I expect there are a lot of users that do not read and compare the charts carefully but simply go with matching the part number.
I know I did at first and only later really understand what JJ is selling is more of a 7591 "inspired" tube that may work in some applications.
Looking at the tunsol data sheet you are correct that the match is better with the Electroharmonix tube than the JJ tube.
The tunsol is still significantly weaker than a NOS 7591 but closer than the JJ.