Skimmed both. I'll have to take a deeper dive when I have time, but it looks like you're misreading the first paper, while the second paper is seemingly an outlier in not finding a main effect for attractiveness on mating success (the "girth" finding is consistent with previous literature though).

First, excerpts from the first paper you seem to have overlooked:

We found a small, but significant, correlation between testosterone and number of sexual partners, when controlling for age (t107 = 0.11, P < 0.05).

---

Attractiveness and masculinity were significantly correlated with each other (partial r104 = 0.75, P < 0.001), and both were significantly correlated with mating success, controlling for age (attractiveness: partial t112 = 0.13, P < 0.05; masculinity: partial t112 = 0.17, P < 0.01). Importantly, however, testosterone (controlling for lifestyle factors) was not correlated with either attractiveness (r104 = 0.12, P = 0.22) or masculinity (partial r104 = 0.01, P = 0.94).

---

Despite replicating the finding by Rhodes et al. (2005) that attractive and masculine males have higher mating success, we found no evidence that females were sensitive to, or preferred, visual cues of circulating testosterone in the male face and body, because testosterone did not significantly correlate with either masculinity or attractiveness.

---

---

---

Note the higher coefficients for attractiveness and masculinity (which were themselves highly correlated) when it comes to mating success. Note they confirm they replicated earlier findings they previously published -- here's that paper: https://www.sciencedirect.com/science/article/pii/S1090513804000765

Attractiveness and sexual behavior: Does attractiveness enhance mating success?

Abstract

If attractiveness is an important cue for mate choice, as proposed by evolutionary psychologists, then attractive individuals should have greater mating success than their peers. We tested this hypothesis in a large sample of adults. Facial attractiveness correlated with the number of short-term, but not long-term, sexual partners, for males, and with the number of long-term, but not short-term, sexual partners and age of first sex, for females. Body attractiveness also correlated significantly with the number of short-term, but not long-term, sexual partners, for males, and attractive males became sexually active earlier than their peers. Body attractiveness did not correlate with any sexual behavior variable for females. To determine which aspects of attractiveness were important, we examined associations between sexual behaviors and three components of attractiveness: sexual dimorphism, averageness, and symmetry. Sexual dimorphism showed the clearest associations with sexual behaviors. Masculine males (bodies, similar trend for faces) had more short-term sexual partners, and feminine females (faces) had more long-term sexual partners than their peers. Feminine females (faces) also became sexually active earlier than their peers. Average males (faces and bodies) had more short-term sexual partners and more extra-pair copulations (EPC) than their peers. Symmetric women (faces) became sexually active earlier than their peers. Given that male reproductive success depends more on short-term mating opportunities than does female reproductive success, these findings suggest that individuals of high phenotypic quality have higher mating success than their lower quality counterparts.

---

I hope it's clear crux of Peters, Simmons & Rhodes (2008) is NOT that physical attractiveness is negligible, but instead, that it does not mediate the (weak) relationship between salivary T and mating success. It should also be noted they go out of their way to highlight the limitations of inferring too much from the correlation, given the well-known reverse causality involved in T and sexual activity (i.e., proceptive behavior and having sex raises men's T).

Other relevant considerations (from the following review:https://snelab.nipissingu.ca/wp-content/uploads/sites/32/2014/10/Gray-et-al.pdf ):

Studies of men's testosterone and sexual behavior suggest either weak or null relationships, although other factors qualify these general observations. In a study of 11 couples, men's blood testosterone sampled regularly in a lab was not related to intercourse frequency across a span of three months (Persky et al., 1978). Among 33 young German men, blood testosterone levels sampled six times over a two-week span were positively associated with sexual experiences occurring around the days of sample collection (Knussmann et al., 1986). In a study of 119 young Australian male university students, men's testosterone was weakly but positively correlated with cumulative sexual experience (Peters et al., 2008). However, a different study of approximately 100 young American men found no relationship between men's recent sexual behavior and blood testosterone levels (Brown et al., 1978). In a study of Hadza hunter-gatherer and Datoga pastoralist androgen receptor polymorphisms and male reproductive success, shorter androgen receptor CAG repeats (indicative of greater “androgenicity”) predicted more children, with this link mediated by self-reported aggression on a questionnaire (Butovskaya et al., 2015).

*Should also be noted, for better or worse, that Peters (2008) used quite a bit of statistical manipulation to "control" for "lifestyle factors" purported to influence T, and that without this intervention their T-mating success correlation was no longer significant.

---

---

---

With the second paper, one wonders to what extent selecting participants from two Penn State fraternities -- who collectively exhibit more phenotypic similarity and more narrow variation than what you might find randomly selecting men on campus -- might have made the study less sensitive to detecting the influence of attractiveness.

Finally, it should be emphasized neither of the papers you link easily reconcile their findings with female choice in this process, with the second paper proposing multiple highly speculative hypotheses to explain their findings. That is to say, it is easier to imagine dominance/T-associated behaviors more readily manifesting as increased mating effort / higher aggression / more forgiving mate standards (for short-term mating) by men, than a mere attraction to certain behaviors over physical appearance by women (although these need not be mutually exclusive).

---

tl;dr: interesting Red Pills, one study reaffirms the salience of looks, the other is best characterized as an outlier.