Under the assumption that all of the R values are greater than 0, and all of the C values are greater than 0, and s is greater than zero, then you can demonstrate that the back-substitutions are always true -- the sum of strictly positive values must be greater than 0.
If, however, any of your C values are zero, or s is zero, then you have division by 0 in your equations, and that becomes unsolvable.
syms R0 R1 R2 R3 R4 R5 C1 C2 C3 Vin I1 I2 I3 I4 Vout s H
eq1 = Vin == I1*R0+I1/(C2*s)-I2/(C2*s)+I1*R3-I3*R3+I1/(C1*s)-I4/(C1*s);
eq2 = 0 == I2*R1+I2/(C3*s)-I4/(C3*s)+I2*R4-I3*R4+I2/(C2*s)-I1/(C2*s);
eq3 = 0 == I3*R3-I1*R3+I3*R4-I2*R4+I3*R5-I4*R5;
eq4 = 0 == I4*R2+I4/(C1*s)-I1/(C1*s)+I4*R5-I3*R5+I4/(C3*s)-I2/(C3*s);
result = solve([eq1,eq2,eq3,eq4,eq5,eq6],[Vin,I1,I2,I3,I4,Vout,H],ReturnConditions=false);
backsubs = subs([eq1,eq2,eq3,eq4,eq5,eq6], result)