No. Lonchampt referred to both. He noticed that the denominator of the relation [3] of his paper (1), that is the difference (P*-P) between the atmospheric pressure (P*) and saturation pressure (P) at the cell temperature (that he erroneously called "the water pressure at the temperature of the bath"), goes to zero as cell temperature approaches the boiling value, therefore this relationship can't be used without introducing huge errors.
Yes, Lonchampt has wisely choosen to calculate the energy balance in the boiling region by using his relation [6], which is equivalent to the calculations on Page 16 of the F&P paper (2). But unlike F&P, who arbitrarily and erroneously applied this method to the supposed last 10 minutes of boil-off, Lonchampt applied this method starting from the moment when "temperature reaches a value close to boiling, i.e. typically 99 to 101°C", that is for the whole period of many hours during which the water in the cell was boiling.
However this Lonchampt's criterion to mark the onset of boiling is still inadequate: the boiling on the electrode surfaces begins when the water bulk temperature is lower than 99°C. A better criterion would have been to start from the moment when the input power becomes greater that the 9-11 W estimated as the heat losses by radiation/conduction.
In each diagram shown in the Lonchampt paper (1), the label "Excess Heat" must be changed to "Energy Unbalance" or, even better, "Balance Error". The same holds for his relations [1] and [6].
The only thing F&P proved in their Fig.7C of (2) (I guess you are referring to this specific diagram) is their inability to make an energy balance!
Fig.7C is wrong with many respects, starting from erroneous indication of the cell dryness at the time of about 1656 ks, despite the lab video indicates a time of around 1655 ks, that is about 9000 s after the time indicated in Fig.7C.
Fig.7C shows the values of (k'R)11 calculated by the following relation [4] of (2).
The numerator contains 3 terms:
1 - the energy input due to electrolysis;
2 - the energy content of the gas stream, including the heat carried away by the steam which is supposed to saturate the gas bubbles produced by electrolysis;
3 - the change in the enthalpy content of the calorimeter due to the increase of its temperature.
Looking more deeply at the second term, the presence of the difference between the specific heat of gaseous and liquid D2O is questionable, in fact Lonchampt didn't include it in his equivalent relation [3] of (1). But this is a minor issue.
The big problem is that the second term is evidently inadequate to account for the energy losses due to evaporation during boiling, for the following two reasons:
a - the factor P/(P*-P) goes to infinite (as already said);
b - and, most importantly, this term only accounts for the vapor which saturates the gas bubble produced by electrolyses, it doesn't account for the vapor carried out by the bubbles directly produced by boiling!
In other words, in their major paper (2) - which, as stated in the abstract, "concerned with high rates of specific excess enthalpy generation (> 1kWcm-3) at temperatures close to (or at) the boiling point of the electrolyte solution" - F&P incredibly omitted to consider, in their calorimeter model, the enthalpy loss due to vapor produced by boiling!
(1) http://www.lenr-canr.org/acrobat/LonchamptGreproducti.pdf
(2) http://www.lenr-canr.org/acrobat/Fleischmancalorimetra.pdf