R: AdehabitatLT test Correlated Random Walk models (TestNM) taking a long time to compute (A Bug?)

I am trying to create correlated random walk models that are bounded within polygons (100% MCP home ranges), using the NMs.randomCRW function from the adehabitatLT package. I noticed a similar topic "How to Bound Correlated Random Walk Models within a Polygon using R", but our issues appear to be different.

My NMs.randomCRW models are created no problem. When I use the TestNM function to test the NMs models, it only works effectively when no functions are in my model (quickly each time). When I test my null models with a constraint function (constrained to a polygon), it only appears to 'work' <2% of the time (and so far, for only one animal). Most of the time, the TestNM model runs for several hours before I stop the function (figuring it wasn't working properly)... the <2% of the time it works, I get exactly what I want within 2 minutes (or less), making me suspect that my code is correct. (Image here shows randomCRW constrained to polygon (left), actual trajectory in polygon (right): Constrained randomCRW

Does anyone know what my issue may be? Is this a bug with the TestNM function? How does the testNM work? Does it keep running until it 'finds a model' that fits my Null Model?

My code is as follows:

library(adehabitatLT)
library(dplyr)
library(sp)

## Load data and create PFTraj
fc<-read.csv("C:/Users/Brennan/Documents/DataManagement/AnimalswithENV.csv")
da <- as.character(fc$TIMESTAMP_)
dapf <- as.POSIXct(da)
da$dapf <- dapf
PFTraj <- adehabitatLT::as.ltraj(xy = fc[, c("Long_", "Lat")], date=dapf, id=fc$tag_ident)

## Create MCP home Range, 100%
PFs <- structure(list(longitude = fc$Long_, latitude = fc$Lat, BirdID = fc$tag_ident, DateTime = fc$TIMESTAMP_), .Names = c('longitude', 'latitude', 'BirdID', 'DateTime'), class="data.frame", row.names= c(NA, -93382L))
xy <- PFs[,c(1,2)]
PFs2 <- PFs[,c(3,4)]
PFsp <- SpatialPointsDataFrame(coords=xy, data= PFs2, proj4string = CRS("+proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0"))

cp2 <- adehabitatHR::mcp(PFsp[,1], percent = 100)

#Use a consistent projection
proj4string = CRS("+proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0")

## My Constraint Function
confun <- function(x, par) {
  coordinates(x) <- x[,1:2]
  proj4string(x) <- proj4string(par)
  ov <- over(x, par)
  return(all(!is.na(ov)))  
}

#Extract 1 bird from PFTraj and SpatialPolygonsDataFrame (~2000 relocations)
eA1 <- PFTraj[1]
Bird1 <- cp2[cp2@data$id=="3117",]
Bird1 <- as(Bird1, "SpatialPolygons")

## Create Null model and test it
xo1 <- coordinates(Bird1)
nmo1B <- NMs.randomCRW(na.omit(eA1), rangles = TRUE, rdist = TRUE, fixedStart = TRUE,
                       rx = range(xo1[,1]), ry = range(xo1[,2]), treatment.func = NULL, 
                       treatment.par = NULL, constraint.func = confun, constraint.par = Bird1, 
                       nrep = 1)
resu <- testNM(nmo1B, count = FALSE)

## Plot results
resul <- data.frame(resu)
spresu<- SpatialPointsDataFrame(coords=resul[,c(1,2)], data=resul, proj4string = CRS("+proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0"))
plot(Bird1)
plot(spresu, add=TRUE)

Again, this code HAS produced the randomCRW I want, just not very often, and only for one animal so far. HOWEVER, when I replicate this code for my 2nd animal, the TestNM function doesn't complete, even after hours of the test running...

Any help would be appreciated,

Thanks in advance!