Remove closure allocations in two methods

[rogerioyuuki]

Hi, I was benchmarking a few caching libraries and noticed two closure allocations which can be avoided.

1. FastCache.GetOrAdd was using the TArg overload of ConcurrentDictionary (as pointed out by A few small issues #1 and fixed by ebef51e), but it was still capturing valueFactory variable. I changed TArg to a tuple of the two variables needed.
2. FastCache.AddOrUpdate could be similarly changed to use another overload and prevent capturing ttlValue variable.

I ran the project benchmark on my machine with the following results:

BenchmarkDotNet=v0.13.2, OS=Windows 11 (10.0.22621.1702)
12th Gen Intel Core i7-12800H, 1 CPU, 20 logical and 14 physical cores
.NET SDK=7.0.202
  [Host]   : .NET 6.0.14 (6.0.1423.7309), X64 RyuJIT AVX2
  ShortRun : .NET 6.0.14 (6.0.1423.7309), X64 RyuJIT AVX2

Job=ShortRun  IterationCount=3  LaunchCount=1  
WarmupCount=3  

Main branch (3e9a35a)

Method	Mean	Error	StdDev	Gen0	Allocated
DictionaryLookup	57.49 ns	11.028 ns	0.604 ns	-	-
FastCacheLookup	55.94 ns	5.386 ns	0.295 ns	-	-
MemoryCacheLookup	347.25 ns	135.332 ns	7.418 ns	0.0100	128 B
FastCacheGetOrAdd	81.88 ns	12.146 ns	0.666 ns	0.0076	96 B
FastCacheAddRemove	73.22 ns	21.811 ns	1.196 ns	0.0134	168 B
MemoryCacheAddRemove	389.02 ns	80.734 ns	4.425 ns	0.0257	328 B

This PR (686fd04)

Method	Mean	Error	StdDev	Gen0	Allocated
DictionaryLookup	57.67 ns	5.942 ns	0.326 ns	-	-
FastCacheLookup	59.16 ns	25.778 ns	1.413 ns	-	-
MemoryCacheLookup	343.65 ns	145.375 ns	7.969 ns	0.0100	128 B
FastCacheGetOrAdd	71.71 ns	23.982 ns	1.315 ns	-	-
FastCacheAddRemove	72.49 ns	103.096 ns	5.651 ns	0.0063	80 B
MemoryCacheAddRemove	369.58 ns	83.251 ns	4.563 ns	0.0257	328 B

[alex-jitbit]

Thanks!