They did with the P226 when the Beretta 92F/M9 won. The cost of the gun was less than the Beretta 92F IIRC but Beretta won because of the overall total cost.
The cost of the P320 is much lower than the P226. The P320 in 9mm sells for less than $450 all the time to the general public.
****************************************
Of course the 320 is a lower price point than the 226. As for why the 226 wasn't chosen instead of the M9, there's quite a bit more to it than that and a lot of it is political and contract asskissing by brass in certain places.
1) The Sig scored slightly higher than the Beretta.
2) Beretta dropped it's price 18% at the last minute, presumably because it may have obtained leaked price info from the Sig bid. Originally the total cost for the Beretta contract was $84 Million, and the Sig bid was $75 Million.
3) Only the Sig and Beretta passed the technical and reliability standards. The HK P7M13, S&W 459M, Walther P88, Colt SSP, FN-Herstal BDA, and Steyr GB all failed or were withdrawn voluntarily.
4) The Sig pistol was cheaper even after the 18% price drop. The difference in cost on the final contract was only $3 Million, all in the cost of magazines and spare parts.
5) The choice to adopt 9mm was solely on the basis of NATO and domestic caliber standardization. Congress denied funding for a continuation of the M1911A1 and any .45 ACP related programs to force the JSO to make this decision.
6) Testing consisted of a field trial of 275,000 rounds done to test the performance of 30 guns (sometimes as few as 2 of each make) in dust, mud, sand, and salt water. Further testing included temperature variables (hot / cold) and testing by personnel of differing size, gender, and experience.
7) The Army stated quite clearly that .38 caliber revolvers; "have inadequate overall effectiveness; poor maintainability and life expectancy in combat conditions, low-lethality, poor reliability, lack of rapid reloading ability, and small ammunition capacity."
8) HK's P7M13 lowest score on the conditional reliability testing was 99%. It blew the doors off of the Sig P226 and Beretta 92. It scored a 100% on the dust test.
9) The Sig suffered in the dry mud environment, but scored more than 50% higher on the standard reliability testing than the Beretta. The Sig P226 scored a 2,877 to the Beretta's 1,750 to the Colt M1911A1s 162. The numbers reflect the average number of rounds fired between stoppages.
10) Two of the five pistols submitted in 1981 for reliability testing by Beretta accounted for two-thirds of the malfunctions.
11) Reliability between each lot of guns and the individual guns in the lot varied on a factor of 300 - 1400%; that's just comparing Sig to Sig, Beretta to Beretta.
12) The Sig P226 originally in 1981 had a significant issue with the shape of it's firing pin causing a large number of failures to fire. It was reshaped (sharpened) for the 1984 trial and eliminated the issue.
13) You can fire approximately 6,253 rounds through a Sig P226 before the frame cracks with no maintenance. In comparison, the M1911A1 control frame failed at around 3,500 rounds.
14) Sig originally had no frame failures in the 7,000 round testing in 1981. The frame was milled out to improve mud and debris performance, and the resulting hollow sections are where the cracks propagated. Thus, pre-1981 Sigs are certainly more reliable in a long term environment.
15) In the 1984 reliability testing, the Sig P226 had only one malfunction that required an armorer and finished with a total of 12 stoppages. The Beretta had 20 stoppages, with 9 of them requiring an armorer. Thus, you have a near 50% chance that within 1,700 rounds your Beretta 92 will suffer a complete malfunction and require significant maintenance. I can attest from personal experience that this is true of the commercial 92FS from a hard use perspective. The predominance is frame cracking.
16) The M1911A1 had a mean round count of 162 between stoppages, and had a total of 220 stoppages; only 25 required an armorer. I suspect frame cracking was a large part of the issue with the M1911 due to their reliance upon the recoil spring to provide sufficient dampening to the slide during the recoil phase.
17) The Sig P226's firing pin invariably caused roughly 70% of the failures in the 1981 trial.
18) Beretta USA at the time purchased the pistols from Italy for $178.50, and sold them at a retail price of $515.
19) There was significant posturing and lobbying from the Italian government in favor of awarding the contract to Beretta. It was seen that this could be averted by awarding the contract to Beretta USA on the notion that eventually the pistols would be made in Accucreek, MD.
20) The trials were done half-assed and with numerous holes in the methods of testing. The Army and Air Force did not have a standardized method for testing the guns, and relied upon "better than M1911" performance standards as the basis of the testing. As a result, some of the testing numbers were overlooked such as the Sig dry mud testing due to testing methodology shortcomings.
21) Sig sued the US on the basis that the number of replacement parts was not aligned with reliability figures. The significantly more reliable P226 required no replacement parts in 5,000 rounds, yet the pricing quotations stipulated a set number of parts, and in some instances those parts were double counted due to being difficult to field fit, small in size (easily lost), and other variables. The trial was thrown out.
In the end, the US Army should have awarded Sig (SACO) the contract on the basis that it was a more reliable and more cost effective gun. It would seem that the SEALs chose the wiser option in the long run.